Compare commits

...

80 Commits

Author SHA1 Message Date
junos 777d30365e Fix questions that were slightly different in the morning. 2023-07-03 21:29:09 +02:00
junos ca85131ed2 Use refactored methods. 2023-07-03 21:18:15 +02:00
junos 2aa5c5cb07 Fix trailing whitespace. 2023-07-03 21:17:40 +02:00
junos 82b53bc0d3 Extract method to reuse. 2023-07-03 21:13:50 +02:00
junos c688580fe8 Increment answers separately if needed. 2023-07-03 21:01:15 +02:00
junos 8c0b66eddc Extract method to reuse and simplify. 2023-07-03 20:52:08 +02:00
junos e3ff4846e1 Document functions. 2023-07-03 20:45:39 +02:00
junos 825380a47e Add a function to fix SAM question IDs. 2023-07-03 20:41:48 +02:00
junos ef26772038 Add SAM question IDs. 2023-07-03 20:38:20 +02:00
junos 64c05ec5ec Fix COPE question IDs. 2023-07-03 20:27:33 +02:00
junos c0236b251c Fix question IDs. 2023-07-03 20:11:11 +02:00
junos 2f22f2052a [WIP] Continue factor analysis. 2023-07-03 20:10:45 +02:00
junos ec51d7d406 [WIP] Add a function to recode question IDs. 2023-07-03 19:57:28 +02:00
junos 2aca64aa09 Add COPE questions and their IDs. 2023-07-03 19:34:11 +02:00
junos 9a87a0a34a Add instructions to export. 2023-07-03 19:19:10 +02:00
junos 91a9f20839 Read in data. 2023-07-03 18:44:45 +02:00
junos e3be17e56e Export COPE data. 2023-07-03 18:33:28 +02:00
junos 5c0e2e2621 Export SAM data. 2023-07-03 18:25:40 +02:00
junos 51201c2bc9 Disable jupytext pre-commit again. 2023-07-03 18:24:37 +02:00
junos fed4b33611 Add more jupytext args. 2023-07-03 18:22:36 +02:00
junos 471ce7c2cb Document extract_stressful_events. 2023-07-03 17:37:34 +02:00
junos dbb2033f78 Add jupytext to pre-commit hooks. 2023-07-03 17:35:45 +02:00
junos 1b77fb119c Fix an error introduced in ae2ca63bc4. 2023-07-03 17:17:56 +02:00
junos ae2ca63bc4 Define QUESTIONNAIRE IDs and use them.
Clean up docstrings.
2023-07-03 17:09:40 +02:00
junos 577f1330da Add docstrings flake8 checks. 2023-07-03 16:49:35 +02:00
junos 4af360f411 Use conda Python environment with R. 2023-07-03 14:51:07 +02:00
junos 96bbe32f56 Add statistics for some scales. 2023-07-03 14:50:35 +02:00
junos 40170339c2 Add R project and sample Markdown notebook. 2023-07-03 14:28:28 +02:00
junos 7b0c0037f7 Add an old script to test JCQ reversal. 2023-07-03 14:23:55 +02:00
junos db06584ddd Improve removal of "medium" class. 2023-05-31 22:46:49 +02:00
junos 112d968715 Add baseline features. 2023-05-31 22:25:39 +02:00
junos 9cc6bf7c21 Add PCA for composite target. 2023-05-31 21:12:21 +02:00
junos 78807b941c Add analysis for composite score of stress. 2023-05-31 21:00:55 +02:00
junos a9af113c9c Add confusion matrices for all methods. 2023-05-31 17:41:50 +02:00
junos 97113fe9ab Sum up confusion matrix and illustrate use with dummy. 2023-05-31 17:27:49 +02:00
junos bc78a1d498 Define a confusion matrix scorer. 2023-05-31 16:05:39 +02:00
junos aca84b214d Small corrections. 2023-05-19 03:19:42 +02:00
junos aa13123136 Handle clustering classification the same as other classification models again. 2023-05-19 03:04:09 +02:00
junos c51e0da0f7 Handle clustering classification the same as other classification models. 2023-05-19 02:52:56 +02:00
junos a2401b5e36 Add multiclass scoring. 2023-05-19 01:34:34 +02:00
junos 70232949c3 Better handling of "medium" category. 2023-05-19 01:10:30 +02:00
junos 8a9595c615 Better handling of input filename again. 2023-05-18 22:53:58 +02:00
junos 045b9fa0dc Better handling of input filename. 2023-05-18 19:03:53 +02:00
junos bb4445f1a8 Add bins to output filename. 2023-05-18 18:58:19 +02:00
junos 1318ae3609 Rename methods to make them consistent with regression methods. 2023-05-18 18:55:31 +02:00
junos 45441c288d Correct column name. 2023-05-18 18:47:56 +02:00
junos fa45b30955 Set output path programmatically. 2023-05-18 18:40:54 +02:00
junos 2336edffb6 Retain metric names in final scores. 2023-05-18 18:40:06 +02:00
junos b756ed5feb Set more parameters as user-specified constants. 2023-05-18 18:06:32 +02:00
junos cad28c3fe8 Set path programmatically. 2023-05-18 16:36:46 +02:00
junos 38a405d378 Add index when inserting one row. 2023-05-17 18:13:20 +02:00
junos 2c5a0b4157 Label plot axes. 2023-05-17 16:32:27 +02:00
junos 0409c9e982 Fix format specification. 2023-05-16 17:22:09 +02:00
junos a7446cc34a Specify columns to aggregate and save figures as pdfs. 2023-05-16 17:05:43 +02:00
junos 118e686491 Specify format directly as infer_datetime_format was deprecated. 2023-05-16 17:04:48 +02:00
junos 9417a1b9f1 Do not break markdown lines. 2023-05-16 16:37:34 +02:00
junos 7b5db88f1d Remove and ignore results. 2023-05-16 16:22:29 +02:00
junos 0f8f0b0fb6 Update URL call. 2023-05-16 16:17:53 +02:00
junos 26c7d22b83 Add an option to save figures. 2023-05-16 16:17:06 +02:00
junos 87781840d4 Use concat instead of append which was deprecated. 2023-05-12 16:32:08 +02:00
junos 3091328fc5 Format comments. 2023-05-11 16:51:38 +02:00
junos 055e87dbac Return scores for classification. 2023-05-10 23:51:12 +02:00
junos f58d20ffc2 Update classification runner. 2023-05-10 23:17:44 +02:00
junos 075fdab9ea Select segment and save results. 2023-05-10 23:00:03 +02:00
junos 91e7352480 Thoroughly refactor classification runner. 2023-05-10 22:50:00 +02:00
junos 35c09374dd Free up memory during model building. 2023-05-10 21:44:40 +02:00
junos b505fb2b6a Thoroughly refactor regression runner. 2023-05-10 20:30:51 +02:00
junos 47b1ecdbb9 First format with black and then check with flake8. 2023-05-10 15:29:32 +02:00
junos 24744c288d Extract one step of preparation into a separate function. 2023-05-10 15:28:09 +02:00
junos caeaf03239 Provide data instead of csv input. 2023-05-10 15:20:33 +02:00
junos cd5d8b6a10 Update rapids and add regex=True.
Reformat debug_heatmap.
2023-05-10 15:12:27 +02:00
junos 3e38b64b45 Merge branch 'ml_pipeline' 2023-05-10 15:02:17 +02:00
junos 76071fd550 Start using pre-commit hooks. 2023-04-24 15:38:54 +02:00
Primoz 26804cf8ea Repair preprocessing one hot encoding of test set. 2023-04-21 13:24:31 +02:00
Primoz 865225994b Added testing section after feature selection. 2023-04-20 13:29:14 +02:00
Primoz 259be708aa Improve the feature selection method with validations etc. 2023-04-20 13:26:20 +02:00
Primoz 0594993133 Add GroupKFold to feature selection CV. Start with generic metric calculation procedure. 2023-04-20 11:20:26 +02:00
Primoz 1cbc743cf7 Add kBest method to initially filter out the worst performing features. Update comments. 2023-04-20 10:12:16 +02:00
Primoz 2a8f1ee613 Merge branch 'ml_pipeline' of https://repo.ijs.si/junoslukan/straw2analysis into ml_pipeline 2023-04-19 15:56:52 +02:00
Primoz ce13a9e13b Implement feature selection method which is used in ML pipeline. 2023-04-19 15:56:34 +02:00
36 changed files with 2379 additions and 1514 deletions

9
.flake8 100644
View File

@ -0,0 +1,9 @@
[flake8]
max-line-length = 88
extend-ignore =
E203,
# E501 line too long for docstrings
D501
per-file-ignores =
exploration/*.py:E501
docstring-convention = numpy

3
.gitignore vendored
View File

@ -12,12 +12,15 @@ __pycache__/
/data/*input*.csv
/data/daily*
/data/intradaily*
/data/raw
/data/stressfulness_event*
/data/30min*
/presentation/*scores.csv
/presentation/Results.ods
/presentation/results/
.Rproj.user
.Rhistory
/presentation/*.nb.html
presentation/event_stressful_detection_half_loso.csv
presentation/event_stressful_detection_loso.csv
/statistical_analysis/scale_reliability.nb.html

View File

@ -0,0 +1,6 @@
<component name="ProjectCodeStyleConfiguration">
<code_scheme name="Project" version="173">
<option name="RIGHT_MARGIN" value="150" />
<option name="SOFT_MARGINS" value="88" />
</code_scheme>
</component>

View File

@ -0,0 +1,5 @@
<component name="ProjectCodeStyleConfiguration">
<state>
<option name="USE_PER_PROJECT_SETTINGS" value="true" />
</state>
</component>

View File

@ -0,0 +1,3 @@
<component name="ProjectDictionaryState">
<dictionary name="junos" />
</component>

View File

@ -17,7 +17,15 @@
</RMarkdownRenderProfile>
</value>
</entry>
<entry key="file://$PROJECT_DIR$/statistical_analysis/scale_reliability.rmd">
<value>
<RMarkdownRenderProfile>
<option name="lastOutput" value="$PROJECT_DIR$/statistical_analysis/scale_reliability.nb.html" />
<option name="outputDirectoryUrl" value="file://$PROJECT_DIR$/statistical_analysis" />
</RMarkdownRenderProfile>
</value>
</entry>
</map>
</option>
</component>
</project>
</project>

View File

@ -0,0 +1,9 @@
<?xml version="1.0" encoding="UTF-8"?>
<project version="4">
<component name="RGraphicsSettings">
<option name="height" value="600" />
<option name="resolution" value="75" />
<option name="version" value="2" />
<option name="width" value="960" />
</component>
</project>

View File

@ -0,0 +1,7 @@
<?xml version="1.0" encoding="UTF-8"?>
<project version="4">
<component name="RMarkdownGraphicsSettings">
<option name="globalResolution" value="75" />
<option name="version" value="2" />
</component>
</project>

View File

@ -0,0 +1,6 @@
<?xml version="1.0" encoding="UTF-8"?>
<project version="4">
<component name="RSettings">
<option name="interpreterPath" value="C:\Program Files\R\R-4.3.1\bin\R.exe" />
</component>
</project>

View File

@ -0,0 +1,30 @@
repos:
- repo: https://github.com/pre-commit/pre-commit-hooks
rev: v4.4.0
hooks:
- id: check-yaml
- id: end-of-file-fixer
- id: trailing-whitespace
- repo: https://github.com/pycqa/isort
rev: 5.12.0
hooks:
- id: isort
name: isort (python)
- repo: https://github.com/psf/black
rev: 23.3.0
hooks:
- id: black
language_version: python3
- repo: https://github.com/pycqa/flake8
rev: 6.0.0
hooks:
- id: flake8
# - repo: https://github.com/mwouts/jupytext
# rev: v1.14.7
# hooks:
# - id: jupytext
# args: [ --from, "py:percent", --to, "ipynb" ]
# additional_dependencies:
# - isort==5.12.0 # Matches hook
# - black==23.3.0
# - flake8==6.0.0

View File

@ -6,6 +6,7 @@ dependencies:
- black
- isort
- flake8
- flake8-docstrings
- imbalanced-learn=0.10.0
- jupyterlab
- jupytext
@ -14,6 +15,7 @@ dependencies:
- nodejs
- pandas
- psycopg2 >= 2.9.1
- pre-commit
- python-dotenv
- pytz
- pyprojroot
@ -23,4 +25,4 @@ dependencies:
- sqlalchemy
- statsmodels
- tabulate
- xgboost
- xgboost

View File

@ -14,15 +14,9 @@
# ---
# %%
import os, sys
import importlib
import pandas as pd
import numpy as np
# import plotly.graph_objects as go
from importlib import util
from pathlib import Path
import yaml
from rapids.src.features.utils.utils import chunk_episodes
# %%
phone_data_yield = pd.read_csv(
@ -36,23 +30,29 @@ time_segments_labels = pd.read_csv(
# %%
phone_data_yield["assigned_segments"] = phone_data_yield[
"assigned_segments"
].str.replace(r"_RR\d+SS#", "#")
].str.replace(r"_RR\d+SS#", "#", regex=True)
time_segments_labels["label"] = time_segments_labels["label"].str.replace(
r"_RR\d+SS$", ""
r"_RR\d+SS$", "", regex=True
)
# %% tags=[]
def filter_data_by_segment(data, time_segment):
def filter_data_by_segment(data, time_segment_current):
data.dropna(subset=["assigned_segments"], inplace=True)
if data.shape[0] == 0: # data is empty
data["local_segment"] = data["timestamps_segment"] = None
return data
datetime_regex = "[0-9]{4}[\-|\/][0-9]{2}[\-|\/][0-9]{2} [0-9]{2}:[0-9]{2}:[0-9]{2}"
timestamps_regex = "[0-9]{13}"
segment_regex = "\[({}#{},{};{},{})\]".format(
time_segment, datetime_regex, datetime_regex, timestamps_regex, timestamps_regex
datetime_regex = (
r"[0-9]{4}[\-|\/][0-9]{2}[\-|\/][0-9]{2} [0-9]{2}:[0-9]{2}:[0-9]{2}"
)
timestamps_regex = r"[0-9]{13}"
segment_regex = r"\[({}#{},{};{},{})\]".format(
time_segment_current,
datetime_regex,
datetime_regex,
timestamps_regex,
timestamps_regex,
)
data["local_segment"] = data["assigned_segments"].str.extract(
segment_regex, expand=True
@ -147,14 +147,17 @@ def getDataForPlot(phone_data_yield_per_segment):
.fillna(0)
)
# transpose the dataframe per local start datetime of the segment and discard the useless index layer
# transpose the dataframe per local start datetime of the segment
# and discard the useless index layer
phone_data_yield_per_segment = phone_data_yield_per_segment.groupby(
"local_segment_start_datetimes"
)[["minutes_after_segment_start", "sensor"]].apply(
lambda x: x.set_index("minutes_after_segment_start").transpose()
)
phone_data_yield_per_segment.index = phone_data_yield_per_segment.index.get_level_values(
"local_segment_start_datetimes"
phone_data_yield_per_segment.index = (
phone_data_yield_per_segment.index.get_level_values(
"local_segment_start_datetimes"
)
)
return phone_data_yield_per_segment
@ -227,9 +230,13 @@ phone_data_yield_per_segment.tail()
# # A workaround
# %%
phone_data_yield_per_segment["local_segment_start_datetimes", "minutes_after_segment_start"] = phone_data_yield_per_segment[
phone_data_yield_per_segment[
"local_segment_start_datetimes", "minutes_after_segment_start"
] = phone_data_yield_per_segment[
["local_segment_start_datetimes", "minutes_after_segment_start"]
].drop_duplicates(keep="first")
].drop_duplicates(
keep="first"
)
# %%
phone_data_yield_per_segment.set_index(
@ -244,8 +251,9 @@ phone_data_yield_per_segment.head()
# %% [markdown]
# # Retry
# %%
def getDataForPlot(phone_data_yield_per_segment):
def get_data_for_plot(phone_data_yield_per_segment):
# calculate the length (in minute) of per segment instance
phone_data_yield_per_segment["length"] = (
phone_data_yield_per_segment["timestamps_segment"]
@ -292,7 +300,10 @@ def getDataForPlot(phone_data_yield_per_segment):
full_index,
names=("local_segment_start_datetimes", "minutes_after_segment_start"),
)
phone_data_yield_per_segment = phone_data_yield_per_segment.drop_duplicates(subset=["local_segment_start_datetimes", "minutes_after_segment_start"],keep="first")
phone_data_yield_per_segment = phone_data_yield_per_segment.drop_duplicates(
subset=["local_segment_start_datetimes", "minutes_after_segment_start"],
keep="first",
)
phone_data_yield_per_segment = (
phone_data_yield_per_segment.set_index(
["local_segment_start_datetimes", "minutes_after_segment_start"]
@ -302,14 +313,17 @@ def getDataForPlot(phone_data_yield_per_segment):
.fillna(0)
)
# transpose the dataframe per local start datetime of the segment and discard the useless index layer
# transpose the dataframe per local start datetime of the segment
# and discard the useless index layer
phone_data_yield_per_segment = phone_data_yield_per_segment.groupby(
"local_segment_start_datetimes"
)[["minutes_after_segment_start", "sensor"]].apply(
lambda x: x.set_index("minutes_after_segment_start").transpose()
)
phone_data_yield_per_segment.index = phone_data_yield_per_segment.index.get_level_values(
"local_segment_start_datetimes"
phone_data_yield_per_segment.index = (
phone_data_yield_per_segment.index.get_level_values(
"local_segment_start_datetimes"
)
)
return phone_data_yield_per_segment
@ -318,6 +332,6 @@ def getDataForPlot(phone_data_yield_per_segment):
phone_data_yield_per_segment = filter_data_by_segment(phone_data_yield, time_segment)
# %%
data_for_plot_per_segment = getDataForPlot(phone_data_yield_per_segment)
data_for_plot_per_segment = get_data_for_plot(phone_data_yield_per_segment)
# %%

View File

@ -7,7 +7,7 @@
# extension: .py
# format_name: percent
# format_version: '1.3'
# jupytext_version: 1.13.0
# jupytext_version: 1.14.5
# kernelspec:
# display_name: straw2analysis
# language: python
@ -15,19 +15,33 @@
# ---
# %%
import os
import sys
import datetime
import seaborn as sns
nb_dir = os.path.split(os.getcwd())[0]
if nb_dir not in sys.path:
sys.path.append(nb_dir)
import participants.query_db
from features.esm import *
from features.esm_JCQ import *
from features.esm_SAM import *
from features.esm import (
QUESTIONNAIRE_IDS,
clean_up_esm,
get_esm_data,
increment_answers,
preprocess_esm,
reassign_question_ids,
)
from features.esm_COPE import DICT_COPE_QUESTION_IDS
from features.esm_JCQ import reverse_jcq_demand_control_scoring
from features.esm_SAM import DICT_SAM_QUESTION_IDS, extract_stressful_events
# import os
# import sys
# nb_dir = os.path.split(os.getcwd())[0]
# if nb_dir not in sys.path:
# sys.path.append(nb_dir)
# %%
save_figs = False
export_data = True
# %%
participants_inactive_usernames = participants.query_db.get_usernames(
@ -43,8 +57,14 @@ df_esm_preprocessed = preprocess_esm(df_esm_inactive)
# %%
df_esm_PANAS = df_esm_preprocessed[
(df_esm_preprocessed["questionnaire_id"] == 8)
| (df_esm_preprocessed["questionnaire_id"] == 9)
(
df_esm_preprocessed["questionnaire_id"]
== QUESTIONNAIRE_IDS["PANAS_positive_affect"]
)
| (
df_esm_preprocessed["questionnaire_id"]
== QUESTIONNAIRE_IDS["PANAS_negative_affect"]
)
]
df_esm_PANAS_clean = clean_up_esm(df_esm_PANAS)
@ -65,35 +85,47 @@ df_esm_PANAS_daily_means = (
# %%
df_esm_PANAS_summary_participant = (
df_esm_PANAS_daily_means.groupby(["participant_id", "questionnaire_id"])
.agg(["mean", "median", "std"])
.esm_numeric_mean.agg(["mean", "median", "std"])
.reset_index(col_level=1)
)
df_esm_PANAS_summary_participant.columns = df_esm_PANAS_summary_participant.columns.get_level_values(
1
)
df_esm_PANAS_summary_participant[
"PANAS_subscale"
"PANAS subscale"
] = df_esm_PANAS_daily_means.questionnaire_id.astype("category").cat.rename_categories(
{8.0: "PA", 9.0: "NA"}
{8.0: "positive affect", 9.0: "negative affect"}
)
# %%
sns.displot(
data=df_esm_PANAS_summary_participant, x="mean", hue="PANAS_subscale", binwidth=0.2
df_esm_PANAS_summary_participant.groupby("PANAS subscale").describe()["mean"]
# %%
df_esm_PANAS_summary_participant.groupby("PANAS subscale").describe()["std"]
# %%
df_esm_PANAS_summary_participant.query("std == 0")
# %%
fig1 = sns.displot(
data=df_esm_PANAS_summary_participant, x="mean", hue="PANAS subscale", binwidth=0.2
)
fig1.set_axis_labels(x_var="participant mean", y_var="frequency")
if save_figs:
fig1.figure.savefig("PANAS_mean_participant.pdf", dpi=300)
# %%
sns.displot(
data=df_esm_PANAS_summary_participant,
x="median",
hue="PANAS_subscale",
hue="PANAS subscale",
binwidth=0.2,
)
# %%
sns.displot(
data=df_esm_PANAS_summary_participant, x="std", hue="PANAS_subscale", binwidth=0.05
fig2 = sns.displot(
data=df_esm_PANAS_summary_participant, x="std", hue="PANAS subscale", binwidth=0.05
)
fig2.set_axis_labels(x_var="participant standard deviation", y_var="frequency")
if save_figs:
fig2.figure.savefig("PANAS_std_participant.pdf", dpi=300)
# %%
df_esm_PANAS_summary_participant[df_esm_PANAS_summary_participant["std"] < 0.1]
@ -109,8 +141,14 @@ df_SAM_all.head()
# %%
df_esm_SAM = df_esm_preprocessed[
(df_esm_preprocessed["questionnaire_id"] >= 87)
& (df_esm_preprocessed["questionnaire_id"] <= 93)
(
df_esm_preprocessed["questionnaire_id"]
>= QUESTIONNAIRE_IDS["appraisal_stressfulness_event"]
)
& (
df_esm_preprocessed["questionnaire_id"]
<= QUESTIONNAIRE_IDS["appraisal_stressfulness_period"]
)
]
df_esm_SAM_clean = clean_up_esm(df_esm_SAM)
@ -118,9 +156,10 @@ df_esm_SAM_clean = clean_up_esm(df_esm_SAM)
# ## Stressful events
# %%
df_esm_SAM_event = df_esm_SAM_clean[df_esm_SAM_clean["questionnaire_id"] == 87].assign(
stressful_event=lambda x: (x.esm_user_answer_numeric > 0)
)
df_esm_SAM_event = df_esm_SAM_clean[
df_esm_SAM_clean["questionnaire_id"]
== QUESTIONNAIRE_IDS["appraisal_stressfulness_event"]
].assign(stressful_event=lambda x: (x.esm_user_answer_numeric > 0))
# %%
df_esm_SAM_daily_events = (
@ -131,20 +170,22 @@ df_esm_SAM_daily_events = (
)
# %% [markdown]
# Calculate the daily mean of YES (1) or NO (0) answers to the question about a stressful events. This is then the daily ratio of EMA sessions that included a stressful event.
# Calculate the daily mean of YES (1) or NO (0) answers to the question about stressful events. This is then the daily ratio of EMA sessions that included a stressful event.
# %%
df_esm_SAM_event_summary_participant = (
df_esm_SAM_daily_events.groupby(["participant_id"])
.agg(["mean", "median", "std"])
.SAM_event_ratio.agg(["mean", "median", "std"])
.reset_index(col_level=1)
)
df_esm_SAM_event_summary_participant.columns = df_esm_SAM_event_summary_participant.columns.get_level_values(
1
)
# %%
sns.displot(data=df_esm_SAM_event_summary_participant, x="mean", binwidth=0.1)
fig6 = sns.displot(data=df_esm_SAM_event_summary_participant, x="mean", binwidth=0.1)
fig6.set_axis_labels(
x_var="participant proportion of stressful events", y_var="frequency"
)
if save_figs:
fig6.figure.savefig("SAM_events_mean_participant.pdf", dpi=300)
# %%
sns.displot(data=df_esm_SAM_event_summary_participant, x="std", binwidth=0.05)
@ -155,7 +196,12 @@ sns.displot(data=df_esm_SAM_event_summary_participant, x="std", binwidth=0.05)
# %% [markdown]
# * Example of threat: "Did this event make you feel anxious?"
# * Example of challenge: "How eager are you to tackle this event?"
# * Possible answers: 0 - Not at all, 1 - Slightly, 2 - Moderately, 3 - Considerably, 4 - Extremely
# * Possible answers:
# 0 - Not at all,
# 1 - Slightly,
# 2 - Moderately,
# 3 - Considerably,
# 4 - Extremely
# %%
df_esm_SAM_daily = (
@ -167,27 +213,45 @@ df_esm_SAM_daily = (
# %%
df_esm_SAM_daily_threat_challenge = df_esm_SAM_daily[
(df_esm_SAM_daily["questionnaire_id"] == 88)
| (df_esm_SAM_daily["questionnaire_id"] == 89)
(df_esm_SAM_daily["questionnaire_id"] == QUESTIONNAIRE_IDS["appraisal_threat"])
| (df_esm_SAM_daily["questionnaire_id"] == QUESTIONNAIRE_IDS["appraisal_challenge"])
]
# %%
df_esm_SAM_summary_participant = (
df_esm_SAM_daily.groupby(["participant_id", "questionnaire_id"])
.agg(["mean", "median", "std"])
.esm_numeric_mean.agg(["mean", "median", "std"])
.reset_index(col_level=1)
)
df_esm_SAM_summary_participant.columns = df_esm_SAM_summary_participant.columns.get_level_values(
1
# %%
df_esm_SAM_event_stressfulness_summary_participant = df_esm_SAM_summary_participant[
df_esm_SAM_summary_participant["questionnaire_id"]
== QUESTIONNAIRE_IDS["appraisal_stressfulness_event"]
]
df_esm_SAM_event_stressfulness_summary_participant.describe()["mean"]
# %%
df_esm_SAM_event_stressfulness_summary_participant.describe()["std"]
# %%
sns.displot(
data=df_esm_SAM_event_stressfulness_summary_participant, x="mean", binwidth=0.2
)
# %%
df_esm_SAM_threat_challenge_summary_participant = df_esm_SAM_summary_participant[
(df_esm_SAM_summary_participant["questionnaire_id"] == 88)
| (df_esm_SAM_summary_participant["questionnaire_id"] == 89)
(
df_esm_SAM_summary_participant["questionnaire_id"]
== QUESTIONNAIRE_IDS["appraisal_threat"]
)
| (
df_esm_SAM_summary_participant["questionnaire_id"]
== QUESTIONNAIRE_IDS["appraisal_challenge"]
)
]
df_esm_SAM_threat_challenge_summary_participant[
"event_subscale"
"event subscale"
] = df_esm_SAM_threat_challenge_summary_participant.questionnaire_id.astype(
"category"
).cat.rename_categories(
@ -198,26 +262,84 @@ df_esm_SAM_threat_challenge_summary_participant[
sns.displot(
data=df_esm_SAM_threat_challenge_summary_participant,
x="mean",
hue="event_subscale",
hue="event subscale",
binwidth=0.2,
)
# %%
sns.displot(
fig3 = sns.displot(
data=df_esm_SAM_threat_challenge_summary_participant,
x="std",
hue="event_subscale",
hue="event subscale",
binwidth=0.1,
)
fig3.set_axis_labels(x_var="participant standard deviation", y_var="frequency")
if save_figs:
fig3.figure.savefig("SAM_std_participant.pdf", dpi=300)
# %%
df_esm_SAM_threat_challenge_summary_participant.groupby("event subscale").describe()[
"mean"
]
# %%
df_esm_SAM_threat_challenge_summary_participant.groupby("event subscale").describe()[
"std"
]
# %%
df_esm_SAM_clean.columns
# %%
df_esm_SAM_clean.esm_status.value_counts()
# %%
if export_data:
df_esm_SAM_fixed = reassign_question_ids(df_esm_SAM_clean, DICT_SAM_QUESTION_IDS)
df_esm_SAM_fixed = increment_answers(df_esm_SAM_fixed)
df_esm_SAM_for_export = df_esm_SAM_fixed[
[
"participant_id",
"username",
"device_id",
"_id",
"esm_trigger",
"esm_session",
"esm_notification_id",
"question_id",
"questionnaire_id",
"esm_instructions",
"double_esm_user_answer_timestamp",
"datetime_lj",
"date_lj",
"time",
"esm_user_answer",
"esm_user_answer_numeric",
]
]
df_esm_SAM_for_export.sort_values(
by=["participant_id", "device_id", "_id"], ignore_index=True, inplace=True
)
print(df_esm_SAM_for_export.head())
df_esm_SAM_for_export.to_csv(
"../data/raw/df_esm_SAM_threat_challenge.csv", index=False
)
# %% [markdown]
# ## Stressfulness of period
# %%
df_esm_SAM_period_summary_participant = df_esm_SAM_summary_participant[
df_esm_SAM_summary_participant["questionnaire_id"] == 93
df_esm_SAM_summary_participant["questionnaire_id"]
== QUESTIONNAIRE_IDS["appraisal_stressfulness_period"]
]
# %%
df_esm_SAM_period_summary_participant.describe()["mean"]
# %%
df_esm_SAM_period_summary_participant.describe()["std"]
# %%
sns.displot(data=df_esm_SAM_period_summary_participant, x="mean", binwidth=0.2)
@ -229,8 +351,8 @@ sns.displot(data=df_esm_SAM_period_summary_participant, x="std", binwidth=0.1)
# %%
df_esm_JCQ_demand_control = df_esm_preprocessed[
(df_esm_preprocessed["questionnaire_id"] >= 10)
& (df_esm_preprocessed["questionnaire_id"] <= 11)
(df_esm_preprocessed["questionnaire_id"] >= QUESTIONNAIRE_IDS["JCQ_job_demand"])
& (df_esm_preprocessed["questionnaire_id"] <= QUESTIONNAIRE_IDS["JCQ_job_control"])
]
df_esm_JCQ_demand_control_clean = clean_up_esm(df_esm_JCQ_demand_control)
@ -250,14 +372,11 @@ df_esm_JCQ_daily = (
)
df_esm_JCQ_summary_participant = (
df_esm_JCQ_daily.groupby(["participant_id", "questionnaire_id"])
.agg(["mean", "median", "std"])
.esm_score_mean.agg(["mean", "median", "std"])
.reset_index(col_level=1)
)
df_esm_JCQ_summary_participant.columns = df_esm_JCQ_summary_participant.columns.get_level_values(
1
)
df_esm_JCQ_summary_participant[
"JCQ_subscale"
"JCQ subscale"
] = df_esm_JCQ_summary_participant.questionnaire_id.astype(
"category"
).cat.rename_categories(
@ -265,11 +384,71 @@ df_esm_JCQ_summary_participant[
)
# %%
sns.displot(
data=df_esm_JCQ_summary_participant, x="mean", hue="JCQ_subscale", binwidth=0.1,
)
df_esm_JCQ_summary_participant.groupby("JCQ subscale").describe()["mean"]
# %%
sns.displot(
data=df_esm_JCQ_summary_participant, x="std", hue="JCQ_subscale", binwidth=0.05,
df_esm_JCQ_summary_participant.groupby("JCQ subscale").describe()["std"]
# %%
fig4 = sns.displot(
data=df_esm_JCQ_summary_participant,
x="mean",
hue="JCQ subscale",
binwidth=0.1,
)
fig4.set_axis_labels(x_var="participant mean", y_var="frequency")
if save_figs:
fig4.figure.savefig("JCQ_mean_participant.pdf", dpi=300)
# %%
fig5 = sns.displot(
data=df_esm_JCQ_summary_participant,
x="std",
hue="JCQ subscale",
binwidth=0.05,
)
fig6.set_axis_labels(x_var="participant standard deviation", y_var="frequency")
if save_figs:
fig5.figure.savefig("JCQ_std_participant.pdf", dpi=300)
# %% [markdown]
# # COPE Inventory
# %%
df_esm_COPE = df_esm_preprocessed[
(df_esm_preprocessed["questionnaire_id"] >= QUESTIONNAIRE_IDS["COPE_active"])
& (df_esm_preprocessed["questionnaire_id"] <= QUESTIONNAIRE_IDS["COPE_emotions"])
]
# %%
df_esm_COPE_clean = clean_up_esm(df_esm_COPE)
df_esm_COPE_clean = increment_answers(df_esm_COPE_clean)
df_esm_COPE_fixed = reassign_question_ids(df_esm_COPE_clean, DICT_COPE_QUESTION_IDS)
# %%
if export_data:
df_esm_COPE_for_export = df_esm_COPE_fixed[
[
"participant_id",
"username",
"device_id",
"_id",
"esm_trigger",
"esm_session",
"esm_notification_id",
"question_id",
"questionnaire_id",
"esm_instructions",
"double_esm_user_answer_timestamp",
"datetime_lj",
"date_lj",
"time",
"esm_user_answer",
"esm_user_answer_numeric",
]
]
df_esm_COPE_for_export.sort_values(
by=["participant_id", "device_id", "_id"], ignore_index=True, inplace=True
)
print(df_esm_COPE_for_export.head())
df_esm_COPE_for_export.to_csv("../data/raw/df_esm_COPE.csv", index=False)

View File

@ -20,30 +20,74 @@ import numpy as np
import matplotlib.pyplot as plt
import pandas as pd
from sklearn.ensemble import RandomForestClassifier
from sklearn.metrics import recall_score, f1_score
nb_dir = os.path.split(os.getcwd())[0]
if nb_dir not in sys.path:
sys.path.append(nb_dir)
from machine_learning.cross_validation import CrossValidation
from machine_learning.preprocessing import Preprocessing
from machine_learning.feature_selection import FeatureSelection
# %%
df = pd.read_csv("../data/stressfulness_event_with_speech/input_appraisal_stressfulness_event_mean.csv")
index_columns = ["local_segment", "local_segment_label", "local_segment_start_datetime", "local_segment_end_datetime"]
df.set_index(index_columns, inplace=True)
# Create binary target
bins = [-1, 0, 4] # bins for stressfulness (0-4) target
df['target'], edges = pd.cut(df.target, bins=bins, labels=[0, 1], retbins=True, right=True) #['low', 'medium', 'high']
nan_cols = df.columns[df.isna().any()].tolist()
df[nan_cols] = df[nan_cols].fillna(round(df[nan_cols].median(), 0))
cv = CrossValidation(data=df, cv_method="logo")
categorical_columns = ["gender", "startlanguage", "mostcommonactivity", "homelabel"]
interval_feature_list, other_feature_list = [], []
print(df.columns.tolist())
# %%
for split in cv.get_splits():
train_X, train_y, test_X, test_y = cv.get_train_test_sets(split)
pre = Preprocessing(train_X, train_y, test_X, test_y)
pre.one_hot_encode_train_and_test_sets(categorical_columns)
train_X, train_y, test_X, test_y = pre.get_train_test_sets()
print(train_X.shape, test_X.shape)
# Predict before feature selection
rfc = RandomForestClassifier(n_estimators=10)
rfc.fit(train_X, train_y)
predictions = rfc.predict(test_X)
print("Recall:", recall_score(test_y, predictions))
print("F1:", f1_score(test_y, predictions))
# Feature selection on train set
train_groups, test_groups = cv.get_groups_sets(split)
fs = FeatureSelection(train_X, train_y, train_groups)
selected_features = fs.select_features(n_min=20, n_max=29, k=40,
ml_type="classification_bin",
metric="recall", n_tolerance=20)
train_X = train_X[selected_features]
test_X = test_X[selected_features]
print(selected_features)
print(len(selected_features))
# Predict after feature selection
rfc = RandomForestClassifier(n_estimators=500)
rfc.fit(train_X, train_y)
predictions = rfc.predict(test_X)
print("Recall:", recall_score(test_y, predictions))
print("F1:", f1_score(test_y, predictions))
break
# %%

View File

@ -6,457 +6,129 @@
# extension: .py
# format_name: percent
# format_version: '1.3'
# jupytext_version: 1.13.0
# jupytext_version: 1.14.5
# kernelspec:
# display_name: straw2analysis
# language: python
# name: straw2analysis
# ---
# %% jupyter={"source_hidden": false, "outputs_hidden": false}
# %matplotlib inline
import os
import sys
# %% jupyter={"outputs_hidden": false, "source_hidden": false}
# from IPython.core.interactiveshell import InteractiveShell
from pathlib import Path
import numpy as np
import matplotlib.pyplot as plt
# matplotlib inline
# import os
# import sys
import pandas as pd
from sklearn import linear_model, svm, naive_bayes, neighbors, tree, ensemble
from sklearn.model_selection import LeaveOneGroupOut, cross_validate, StratifiedKFold
from sklearn.dummy import DummyClassifier
from sklearn.impute import SimpleImputer
from machine_learning.helper import (
impute_encode_categorical_features,
prepare_cross_validator,
prepare_sklearn_data_format,
run_all_classification_models,
)
from lightgbm import LGBMClassifier
import xgboost as xg
from IPython.core.interactiveshell import InteractiveShell
InteractiveShell.ast_node_interactivity = "all"
nb_dir = os.path.split(os.getcwd())[0]
if nb_dir not in sys.path:
sys.path.append(nb_dir)
import machine_learning.helper
# %% [markdown]
# # RAPIDS models
# %% [markdown]
# ## Set script's parameters
# InteractiveShell.ast_node_interactivity = "all"
#
# %% jupyter={"source_hidden": false, "outputs_hidden": false} nteract={"transient": {"deleting": false}}
cv_method_str = 'logo' # logo, half_logo, 5kfold # Cross-validation method (could be regarded as a hyperparameter)
n_sl = 3 # Number of largest/smallest accuracies (of particular CV) outputs
undersampling = False # (bool) If True this will train and test data on balanced dataset (using undersampling method)
# %% jupyter={"source_hidden": false, "outputs_hidden": false}
model_input = pd.read_csv("../data/stressfulness_event_with_target_0_ver2/input_appraisal_stressfulness_event_mean.csv")
# model_input = model_input[model_input.columns.drop(list(model_input.filter(regex='empatica_temperature')))]
# %% jupyter={"source_hidden": false, "outputs_hidden": false}
index_columns = ["local_segment", "local_segment_label", "local_segment_start_datetime", "local_segment_end_datetime"]
model_input.set_index(index_columns, inplace=True)
model_input['target'].value_counts()
# %% jupyter={"source_hidden": false, "outputs_hidden": false}
# bins = [-10, 0, 10] # bins for z-scored targets
bins = [-1, 0, 4] # bins for stressfulness (0-4) target
model_input['target'], edges = pd.cut(model_input.target, bins=bins, labels=['low', 'high'], retbins=True, right=True) #['low', 'medium', 'high']
model_input['target'].value_counts(), edges
# model_input = model_input[model_input['target'] != "medium"]
model_input['target'] = model_input['target'].astype(str).apply(lambda x: 0 if x == "low" else 1)
model_input['target'].value_counts()
# %% jupyter={"source_hidden": false, "outputs_hidden": false}
# UnderSampling
if undersampling:
no_stress = model_input[model_input['target'] == 0]
stress = model_input[model_input['target'] == 1]
no_stress = no_stress.sample(n=len(stress))
model_input = pd.concat([stress,no_stress], axis=0)
# model_input_new = pd.DataFrame(columns=model_input.columns)
# for pid in model_input["pid"].unique():
# stress = model_input[(model_input["pid"] == pid) & (model_input['target'] == 1)]
# no_stress = model_input[(model_input["pid"] == pid) & (model_input['target'] == 0)]
# if (len(stress) == 0):
# continue
# if (len(no_stress) == 0):
# continue
# model_input_new = pd.concat([model_input_new, stress], axis=0)
# no_stress = no_stress.sample(n=min(len(stress), len(no_stress)))
# # In case there are more stress samples than no_stress, take all instances of no_stress.
# model_input_new = pd.concat([model_input_new, no_stress], axis=0)
# model_input = model_input_new
# model_input_new = pd.concat([model_input_new, no_stress], axis=0)
# nb_dir = os.path.split(os.getcwd())[0]
# if nb_dir not in sys.path:
# sys.path.append(nb_dir)
# %% jupyter={"source_hidden": false, "outputs_hidden": false}
if cv_method_str == 'half_logo':
model_input['pid_index'] = model_input.groupby('pid').cumcount()
model_input['pid_count'] = model_input.groupby('pid')['pid'].transform('count')
model_input["pid_index"] = (model_input['pid_index'] / model_input['pid_count'] + 1).round()
model_input["pid_half"] = model_input["pid"] + "_" + model_input["pid_index"].astype(int).astype(str)
data_x, data_y, data_groups = model_input.drop(["target", "pid", "pid_index", "pid_half"], axis=1), model_input["target"], model_input["pid_half"]
else:
data_x, data_y, data_groups = model_input.drop(["target", "pid"], axis=1), model_input["target"], model_input["pid"]
# %% jupyter={"source_hidden": false, "outputs_hidden": false}
categorical_feature_colnames = ["gender", "startlanguage"]
additional_categorical_features = [col for col in data_x.columns if "mostcommonactivity" in col or "homelabel" in col]
categorical_feature_colnames += additional_categorical_features
categorical_features = data_x[categorical_feature_colnames].copy()
mode_categorical_features = categorical_features.mode().iloc[0]
# fillna with mode
categorical_features = categorical_features.fillna(mode_categorical_features)
# one-hot encoding
categorical_features = categorical_features.apply(lambda col: col.astype("category"))
if not categorical_features.empty:
categorical_features = pd.get_dummies(categorical_features)
numerical_features = data_x.drop(categorical_feature_colnames, axis=1)
train_x = pd.concat([numerical_features, categorical_features], axis=1)
train_x.dtypes
# %% jupyter={"source_hidden": false, "outputs_hidden": false}
cv_method = StratifiedKFold(n_splits=5, shuffle=True) # Defaults to 5 k-folds in cross_validate method
if cv_method_str == 'logo' or cv_method_str == 'half_logo':
cv_method = LeaveOneGroupOut()
cv_method.get_n_splits(
train_x,
data_y,
groups=data_groups,
)
# %% jupyter={"source_hidden": false, "outputs_hidden": false}
imputer = SimpleImputer(missing_values=np.nan, strategy='median')
# %% [markdown]
# ### Baseline: Dummy Classifier (most frequent)
# %% jupyter={"source_hidden": false, "outputs_hidden": false} nteract={"transient": {"deleting": false}}
dummy_class = DummyClassifier(strategy="most_frequent")
# %% jupyter={"source_hidden": false, "outputs_hidden": false}
dummy_classifier = cross_validate(
dummy_class,
X=imputer.fit_transform(train_x),
y=data_y,
groups=data_groups,
cv=cv_method,
n_jobs=-1,
error_score='raise',
scoring=('accuracy', 'precision', 'recall', 'f1')
)
# %% jupyter={"source_hidden": false, "outputs_hidden": false}
print("Acc (median)", np.nanmedian(dummy_classifier['test_accuracy']))
print("Acc (mean)", np.mean(dummy_classifier['test_accuracy']))
print("Precision", np.mean(dummy_classifier['test_precision']))
print("Recall", np.mean(dummy_classifier['test_recall']))
print("F1", np.mean(dummy_classifier['test_f1']))
print(f"Largest {n_sl} ACC:", np.sort(-np.partition(-dummy_classifier['test_accuracy'], n_sl)[:n_sl])[::-1])
print(f"Smallest {n_sl} ACC:", np.sort(np.partition(dummy_classifier['test_accuracy'], n_sl)[:n_sl]))
# %% [markdown] nteract={"transient": {"deleting": false}}
# ### All models
# %% jupyter={"source_hidden": false, "outputs_hidden": false} nteract={"transient": {"deleting": false}}
final_scores = machine_learning.helper.run_all_classification_models(imputer.fit_transform(train_x), data_y, data_groups, cv_method)
# %% jupyter={"source_hidden": false, "outputs_hidden": false} nteract={"transient": {"deleting": false}}
# %%
final_scores.index.name = "metric"
final_scores = final_scores.set_index(["method", final_scores.index])
final_scores.to_csv(f"../presentation/event_stressful_detection_{cv_method_str}.csv")
CV_METHOD = "logo" # logo, half_logo, 5kfold
# Cross-validation method (could be regarded as a hyperparameter)
print("CV_METHOD: " + CV_METHOD)
N_SL = 3 # Number of largest/smallest accuracies (of particular CV) outputs
UNDERSAMPLING = False
# (bool) If True this will train and test data on balanced dataset
# (using undersampling method)
# %% [markdown]
# ### Logistic Regression
# %% jupyter={"outputs_hidden": false, "source_hidden": false}
PATH_BASE = Path("E:/STRAWresults/20230415")
# %% jupyter={"source_hidden": false, "outputs_hidden": false}
logistic_regression = linear_model.LogisticRegression()
SEGMENT_TYPE = "period"
print("SEGMENT_TYPE: " + SEGMENT_TYPE)
SEGMENT_LENGTH = "30_minutes_before"
print("SEGMENT_LENGTH: " + SEGMENT_LENGTH)
TARGET_VARIABLE = "JCQ_job_control"
print("TARGET_VARIABLE: " + TARGET_VARIABLE)
# %% jupyter={"source_hidden": false, "outputs_hidden": false}
log_reg_scores = cross_validate(
logistic_regression,
X=imputer.fit_transform(train_x),
y=data_y,
groups=data_groups,
cv=cv_method,
n_jobs=-1,
scoring=('accuracy', 'precision', 'recall', 'f1')
if ("appraisal" in TARGET_VARIABLE) and ("stressfulness" in TARGET_VARIABLE):
TARGET_VARIABLE += "_"
TARGET_VARIABLE += SEGMENT_TYPE
PATH_FULL = PATH_BASE / SEGMENT_LENGTH / ("input_" + TARGET_VARIABLE + "_mean.csv")
model_input = pd.read_csv(PATH_FULL)
if SEGMENT_LENGTH == "daily":
DAY_LENGTH = "daily" # or "working"
print(DAY_LENGTH)
model_input = model_input[model_input["local_segment"].str.contains(DAY_LENGTH)]
# %% jupyter={"outputs_hidden": false, "source_hidden": false}
model_input["target"].value_counts()
# %% jupyter={"outputs_hidden": false, "source_hidden": false}
# bins = [-10, 0, 10] # bins for z-scored targets
BINS = [-1, 0, 4] # bins for stressfulness (0-4) target
print("BINS: ", BINS)
model_input["target"], edges = pd.cut(
model_input.target, bins=BINS, labels=["low", "high"], retbins=True, right=True
) # ['low', 'medium', 'high']
print(model_input["target"].value_counts())
REMOVE_MEDIUM = True
if ("medium" in model_input["target"]) and REMOVE_MEDIUM:
model_input = model_input[model_input["target"] != "medium"]
model_input["target"] = (
model_input["target"].astype(str).apply(lambda x: 0 if x == "low" else 1)
)
else:
model_input["target"] = model_input["target"].map(
{"low": 0, "medium": 1, "high": 2}
)
print(model_input["target"].value_counts())
# %% jupyter={"outputs_hidden": false, "source_hidden": false}
# UnderSampling
if UNDERSAMPLING:
no_stress = model_input[model_input["target"] == 0]
stress = model_input[model_input["target"] == 1]
no_stress = no_stress.sample(n=len(stress))
model_input = pd.concat([stress, no_stress], axis=0)
# %% jupyter={"outputs_hidden": false, "source_hidden": false}
model_input_encoded = impute_encode_categorical_features(model_input)
# %%
data_x, data_y, data_groups = prepare_sklearn_data_format(
model_input_encoded, CV_METHOD
)
# %% jupyter={"source_hidden": false, "outputs_hidden": false}
print("Acc (median)", np.nanmedian(log_reg_scores['test_accuracy']))
print("Acc (mean)", np.mean(log_reg_scores['test_accuracy']))
print("Precision", np.mean(log_reg_scores['test_precision']))
print("Recall", np.mean(log_reg_scores['test_recall']))
print("F1", np.mean(log_reg_scores['test_f1']))
print(f"Largest {n_sl} ACC:", np.sort(-np.partition(-log_reg_scores['test_accuracy'], n_sl)[:n_sl])[::-1])
print(f"Smallest {n_sl} ACC:", np.sort(np.partition(log_reg_scores['test_accuracy'], n_sl)[:n_sl]))
cross_validator = prepare_cross_validator(data_x, data_y, data_groups, CV_METHOD)
# %% [markdown]
# ### Support Vector Machine
# %%
data_y.head()
# %% jupyter={"source_hidden": false, "outputs_hidden": false}
svc = svm.SVC()
# %% jupyter={"source_hidden": false, "outputs_hidden": false}
svc_scores = cross_validate(
svc,
X=imputer.fit_transform(train_x),
y=data_y,
groups=data_groups,
cv=cv_method,
n_jobs=-1,
scoring=('accuracy', 'precision', 'recall', 'f1')
# %%
data_y.tail()
# %%
data_y.shape
# %%
scores = run_all_classification_models(data_x, data_y, data_groups, cross_validator)
# %%
PATH_OUTPUT = Path("..") / Path("presentation/results")
path_output_full = PATH_OUTPUT / (
TARGET_VARIABLE
+ "_"
+ SEGMENT_LENGTH
+ "_classification"
+ str(BINS)
+ "_"
+ CV_METHOD
+ ".csv"
)
# %% jupyter={"source_hidden": false, "outputs_hidden": false}
print("Acc (median)", np.nanmedian(svc_scores['test_accuracy']))
print("Acc (mean)", np.mean(svc_scores['test_accuracy']))
print("Precision", np.mean(svc_scores['test_precision']))
print("Recall", np.mean(svc_scores['test_recall']))
print("F1", np.mean(svc_scores['test_f1']))
print(f"Largest {n_sl} ACC:", np.sort(-np.partition(-svc_scores['test_accuracy'], n_sl)[:n_sl])[::-1])
print(f"Smallest {n_sl} ACC:", np.sort(np.partition(svc_scores['test_accuracy'], n_sl)[:n_sl]))
# %% [markdown]
# ### Gaussian Naive Bayes
# %% jupyter={"source_hidden": false, "outputs_hidden": false}
gaussian_nb = naive_bayes.GaussianNB()
# %% jupyter={"source_hidden": false, "outputs_hidden": false}
gaussian_nb_scores = cross_validate(
gaussian_nb,
X=imputer.fit_transform(train_x),
y=data_y,
groups=data_groups,
cv=cv_method,
n_jobs=-1,
error_score='raise',
scoring=('accuracy', 'precision', 'recall', 'f1')
)
# %% jupyter={"source_hidden": false, "outputs_hidden": false}
print("Acc (median)", np.nanmedian(gaussian_nb_scores['test_accuracy']))
print("Acc (mean)", np.mean(gaussian_nb_scores['test_accuracy']))
print("Precision", np.mean(gaussian_nb_scores['test_precision']))
print("Recall", np.mean(gaussian_nb_scores['test_recall']))
print("F1", np.mean(gaussian_nb_scores['test_f1']))
print(f"Largest {n_sl} ACC:", np.sort(-np.partition(-gaussian_nb_scores['test_accuracy'], n_sl)[:n_sl])[::-1])
print(f"Smallest {n_sl} ACC:", np.sort(np.partition(gaussian_nb_scores['test_accuracy'], n_sl)[:n_sl]))
# %% [markdown]
# ### Stochastic Gradient Descent Classifier
# %% jupyter={"source_hidden": false, "outputs_hidden": false}
sgdc = linear_model.SGDClassifier()
# %% jupyter={"source_hidden": false, "outputs_hidden": false}
sgdc_scores = cross_validate(
sgdc,
X=imputer.fit_transform(train_x),
y=data_y,
groups=data_groups,
cv=cv_method,
n_jobs=-1,
error_score='raise',
scoring=('accuracy', 'precision', 'recall', 'f1')
)
# %% jupyter={"source_hidden": false, "outputs_hidden": false}
print("Acc (median)", np.nanmedian(sgdc_scores['test_accuracy']))
print("Acc (mean)", np.mean(sgdc_scores['test_accuracy']))
print("Precision", np.mean(sgdc_scores['test_precision']))
print("Recall", np.mean(sgdc_scores['test_recall']))
print("F1", np.mean(sgdc_scores['test_f1']))
print(f"Largest {n_sl} ACC:", np.sort(-np.partition(-sgdc_scores['test_accuracy'], n_sl)[:n_sl])[::-1])
print(f"Smallest {n_sl} ACC:", np.sort(np.partition(sgdc_scores['test_accuracy'], n_sl)[:n_sl]))
# %% [markdown]
# ### K-nearest neighbors
# %% jupyter={"source_hidden": false, "outputs_hidden": false}
knn = neighbors.KNeighborsClassifier()
# %% jupyter={"source_hidden": false, "outputs_hidden": false}
knn_scores = cross_validate(
knn,
X=imputer.fit_transform(train_x),
y=data_y,
groups=data_groups,
cv=cv_method,
n_jobs=-1,
error_score='raise',
scoring=('accuracy', 'precision', 'recall', 'f1')
)
# %% jupyter={"source_hidden": false, "outputs_hidden": false}
print("Acc (median)", np.nanmedian(knn_scores['test_accuracy']))
print("Acc (mean)", np.mean(knn_scores['test_accuracy']))
print("Precision", np.mean(knn_scores['test_precision']))
print("Recall", np.mean(knn_scores['test_recall']))
print("F1", np.mean(knn_scores['test_f1']))
print(f"Largest {n_sl} ACC:", np.sort(-np.partition(-knn_scores['test_accuracy'], n_sl)[:n_sl])[::-1])
print(f"Smallest {n_sl} ACC:", np.sort(np.partition(knn_scores['test_accuracy'], n_sl)[:n_sl]))
# %% [markdown]
# ### Decision Tree
# %% jupyter={"source_hidden": false, "outputs_hidden": false}
dtree = tree.DecisionTreeClassifier()
# %% jupyter={"source_hidden": false, "outputs_hidden": false}
dtree_scores = cross_validate(
dtree,
X=imputer.fit_transform(train_x),
y=data_y,
groups=data_groups,
cv=cv_method,
n_jobs=-1,
error_score='raise',
scoring=('accuracy', 'precision', 'recall', 'f1')
)
# %% jupyter={"source_hidden": false, "outputs_hidden": false}
print("Acc (median)", np.nanmedian(dtree_scores['test_accuracy']))
print("Acc (mean)", np.mean(dtree_scores['test_accuracy']))
print("Precision", np.mean(dtree_scores['test_precision']))
print("Recall", np.mean(dtree_scores['test_recall']))
print("F1", np.mean(dtree_scores['test_f1']))
print(f"Largest {n_sl} ACC:", np.sort(-np.partition(-dtree_scores['test_accuracy'], n_sl)[:n_sl])[::-1])
print(f"Smallest {n_sl} ACC:", np.sort(np.partition(dtree_scores['test_accuracy'], n_sl)[:n_sl]))
# %% [markdown]
# ### Random Forest Classifier
# %% jupyter={"source_hidden": false, "outputs_hidden": false}
rfc = ensemble.RandomForestClassifier()
# %% jupyter={"source_hidden": false, "outputs_hidden": false}
rfc_scores = cross_validate(
rfc,
X=imputer.fit_transform(train_x),
y=data_y,
groups=data_groups,
cv=cv_method,
n_jobs=-1,
error_score='raise',
scoring=('accuracy', 'precision', 'recall', 'f1'),
return_estimator=True
)
# %% jupyter={"source_hidden": false, "outputs_hidden": false}
print("Acc (median)", np.nanmedian(rfc_scores['test_accuracy']))
print("Acc (mean)", np.mean(rfc_scores['test_accuracy']))
print("Precision", np.mean(rfc_scores['test_precision']))
print("Recall", np.mean(rfc_scores['test_recall']))
print("F1", np.mean(rfc_scores['test_f1']))
print(f"Largest {n_sl} ACC:", np.sort(-np.partition(-rfc_scores['test_accuracy'], n_sl)[:n_sl])[::-1])
print(f"Smallest {n_sl} ACC:", np.sort(np.partition(rfc_scores['test_accuracy'], n_sl)[:n_sl]))
# %% [markdown]
# ### Feature importance (RFC)
# %% jupyter={"source_hidden": false, "outputs_hidden": false}
rfc_es_fimp = pd.DataFrame(columns=list(train_x.columns))
for idx, estimator in enumerate(rfc_scores['estimator']):
feature_importances = pd.DataFrame(estimator.feature_importances_,
index = list(train_x.columns),
columns=['importance'])
# print("\nFeatures sorted by their score for estimator {}:".format(idx))
# print(feature_importances.sort_values('importance', ascending=False).head(10))
rfc_es_fimp = pd.concat([rfc_es_fimp, feature_importances]).groupby(level=0).mean()
pd.set_option('display.max_rows', 100)
print(rfc_es_fimp.sort_values('importance', ascending=False).head(30))
rfc_es_fimp.sort_values('importance', ascending=False).head(30).plot.bar()
rfc_es_fimp.sort_values('importance', ascending=False).tail(30).plot.bar()
train_x['empatica_temperature_cr_stdDev_X_SO_mean'].value_counts()
# %% [markdown]
# ### Gradient Boosting Classifier
# %% jupyter={"source_hidden": false, "outputs_hidden": false}
gbc = ensemble.GradientBoostingClassifier()
# %% jupyter={"source_hidden": false, "outputs_hidden": false}
gbc_scores = cross_validate(
gbc,
X=imputer.fit_transform(train_x),
y=data_y,
groups=data_groups,
cv=cv_method,
n_jobs=-1,
error_score='raise',
scoring=('accuracy', 'precision', 'recall', 'f1')
)
# %% jupyter={"source_hidden": false, "outputs_hidden": false}
print("Acc (median)", np.nanmedian(gbc_scores['test_accuracy']))
print("Acc (mean)", np.mean(gbc_scores['test_accuracy']))
print("Precision", np.mean(gbc_scores['test_precision']))
print("Recall", np.mean(gbc_scores['test_recall']))
print("F1", np.mean(gbc_scores['test_f1']))
print(f"Largest {n_sl} ACC:", np.sort(-np.partition(-gbc_scores['test_accuracy'], n_sl)[:n_sl])[::-1])
print(f"Smallest {n_sl} ACC:", np.sort(np.partition(gbc_scores['test_accuracy'], n_sl)[:n_sl]))
# %% [markdown]
# ### LGBM Classifier
# %% jupyter={"source_hidden": false, "outputs_hidden": false}
lgbm = LGBMClassifier()
# %% jupyter={"source_hidden": false, "outputs_hidden": false}
lgbm_scores = cross_validate(
lgbm,
X=imputer.fit_transform(train_x),
y=data_y,
groups=data_groups,
cv=cv_method,
n_jobs=-1,
error_score='raise',
scoring=('accuracy', 'precision', 'recall', 'f1')
)
# %% jupyter={"source_hidden": false, "outputs_hidden": false}
print("Acc (median)", np.nanmedian(lgbm_scores['test_accuracy']))
print("Acc (mean)", np.mean(lgbm_scores['test_accuracy']))
print("Precision", np.mean(lgbm_scores['test_precision']))
print("Recall", np.mean(lgbm_scores['test_recall']))
print("F1", np.mean(lgbm_scores['test_f1']))
print(f"Largest {n_sl} ACC:", np.sort(-np.partition(-lgbm_scores['test_accuracy'], n_sl)[:n_sl])[::-1])
print(f"Smallest {n_sl} ACC:", np.sort(np.partition(lgbm_scores['test_accuracy'], n_sl)[:n_sl]))
# %% [markdown]
# ### XGBoost Classifier
# %% jupyter={"source_hidden": false, "outputs_hidden": false}
xgb_classifier = xg.sklearn.XGBClassifier()
# %% jupyter={"source_hidden": false, "outputs_hidden": false}
xgb_classifier_scores = cross_validate(
xgb_classifier,
X=imputer.fit_transform(train_x),
y=data_y,
groups=data_groups,
cv=cv_method,
n_jobs=-1,
error_score='raise',
scoring=('accuracy', 'precision', 'recall', 'f1')
)
# %% jupyter={"source_hidden": false, "outputs_hidden": false}
print("Acc (median)", np.nanmedian(xgb_classifier_scores['test_accuracy']))
print("Acc (mean)", np.mean(xgb_classifier_scores['test_accuracy']))
print("Precision", np.mean(xgb_classifier_scores['test_precision']))
print("Recall", np.mean(xgb_classifier_scores['test_recall']))
print("F1", np.mean(xgb_classifier_scores['test_f1']))
print(f"Largest {n_sl} ACC:", np.sort(-np.partition(-xgb_classifier_scores['test_accuracy'], n_sl)[:n_sl])[::-1])
print(f"Smallest {n_sl} ACC:", np.sort(np.partition(xgb_classifier_scores['test_accuracy'], n_sl)[:n_sl]))
scores.to_csv(path_output_full, index=False)

View File

@ -0,0 +1,177 @@
# ---
# jupyter:
# jupytext:
# formats: ipynb,py:percent
# text_representation:
# extension: .py
# format_name: percent
# format_version: '1.3'
# jupytext_version: 1.14.5
# kernelspec:
# display_name: straw2analysis
# language: python
# name: straw2analysis
# ---
# %% jupyter={"outputs_hidden": false, "source_hidden": false}
from pathlib import Path
import pandas as pd
import seaborn as sns
from sklearn.decomposition import PCA
from machine_learning.helper import (
impute_encode_categorical_features,
prepare_cross_validator,
prepare_sklearn_data_format,
run_all_classification_models,
)
# %%
CV_METHOD = "logo" # logo, half_logo, 5kfold
# Cross-validation method (could be regarded as a hyperparameter)
print("CV_METHOD: " + CV_METHOD)
N_SL = 3 # Number of largest/smallest accuracies (of particular CV) outputs
UNDERSAMPLING = False
# (bool) If True this will train and test data on balanced dataset
# (using undersampling method)
# %% jupyter={"outputs_hidden": false, "source_hidden": false}
PATH_BASE = Path("E:/STRAWresults/20230415")
SEGMENT_TYPE = "period"
print("SEGMENT_TYPE: " + SEGMENT_TYPE)
SEGMENT_LENGTH = "30_minutes_before"
print("SEGMENT_LENGTH: " + SEGMENT_LENGTH)
PATH_FULL = PATH_BASE / SEGMENT_LENGTH / "features" / "all_sensor_features.csv"
all_features_with_baseline = pd.read_csv(PATH_FULL)
# %%
TARGETS = [
"PANAS_negative_affect_mean",
"PANAS_positive_affect_mean",
"JCQ_job_demand_mean",
"JCQ_job_control_mean",
"appraisal_stressfulness_period_mean",
]
# %%
all_features_cleaned = pd.DataFrame()
for target in TARGETS:
PATH_FULL = (
PATH_BASE
/ SEGMENT_LENGTH
/ "features"
/ ("all_sensor_features_cleaned_straw_py_(" + target + ").csv")
)
current_features = pd.read_csv(PATH_FULL, index_col="local_segment")
if all_features_cleaned.empty:
all_features_cleaned = current_features
else:
all_features_cleaned = all_features_cleaned.join(
current_features[("phone_esm_straw_" + target)],
how="inner",
rsuffix="_" + target,
)
print(all_features_cleaned.shape)
# %%
pca = PCA(n_components=1)
TARGETS_PREFIXED = ["phone_esm_straw_" + target for target in TARGETS]
pca.fit(all_features_cleaned[TARGETS_PREFIXED])
print(pca.explained_variance_ratio_)
# %%
model_input = all_features_cleaned.drop(columns=TARGETS_PREFIXED)
model_input["target"] = pca.fit_transform(all_features_cleaned[TARGETS_PREFIXED])
# %%
sns.histplot(data=model_input, x="target")
# %%
model_input.target.quantile(0.6)
# %% jupyter={"outputs_hidden": false, "source_hidden": false}
# bins = [-10, 0, 10] # bins for z-scored targets
BINS = [-10, 0, 10] # bins for stressfulness (0-4) target
print("BINS: ", BINS)
model_input["target"], edges = pd.cut(
model_input.target, bins=BINS, labels=["low", "high"], retbins=True, right=True
) # ['low', 'medium', 'high']
print(model_input["target"].value_counts())
REMOVE_MEDIUM = True
if REMOVE_MEDIUM:
if "medium" in model_input["target"]:
model_input = model_input[model_input["target"] != "medium"]
model_input["target"] = (
model_input["target"].astype(str).apply(lambda x: 0 if x == "low" else 1)
)
else:
model_input["target"] = model_input["target"].map(
{"low": 0, "medium": 1, "high": 2}
)
print(model_input["target"].value_counts())
# %% jupyter={"outputs_hidden": false, "source_hidden": false}
# UnderSampling
if UNDERSAMPLING:
no_stress = model_input[model_input["target"] == 0]
stress = model_input[model_input["target"] == 1]
no_stress = no_stress.sample(n=len(stress))
model_input = pd.concat([stress, no_stress], axis=0)
# %%
TARGET_VARIABLE = "PANAS_negative_affect"
print("TARGET_VARIABLE: " + TARGET_VARIABLE)
PATH_FULL_HELP = PATH_BASE / SEGMENT_LENGTH / ("input_" + TARGET_VARIABLE + "_mean.csv")
model_input_with_baseline = pd.read_csv(PATH_FULL_HELP, index_col="local_segment")
# %%
baseline_col_names = [
col for col in model_input_with_baseline.columns if col not in model_input.columns
]
print(baseline_col_names)
# %%
model_input = model_input.join(
model_input_with_baseline[baseline_col_names], how="left"
)
model_input.reset_index(inplace=True)
# %%
model_input_encoded = impute_encode_categorical_features(model_input)
# %%
data_x, data_y, data_groups = prepare_sklearn_data_format(
model_input_encoded, CV_METHOD
)
cross_validator = prepare_cross_validator(data_x, data_y, data_groups, CV_METHOD)
# %%
data_y.head()
# %%
data_y.tail()
# %%
data_y.shape
# %%
scores = run_all_classification_models(data_x, data_y, data_groups, cross_validator)
# %%
PATH_OUTPUT = Path("..") / Path("presentation/results")
path_output_full = PATH_OUTPUT / (
"composite_"
+ SEGMENT_LENGTH
+ "_classification"
+ str(BINS)
+ "_"
+ CV_METHOD
+ ".csv"
)
scores.to_csv(path_output_full, index=False)

View File

@ -6,7 +6,7 @@
# extension: .py
# format_name: percent
# format_version: '1.3'
# jupytext_version: 1.13.0
# jupytext_version: 1.14.5
# kernelspec:
# display_name: straw2analysis
# language: python
@ -14,80 +14,85 @@
# ---
# %% jupyter={"source_hidden": true}
# %matplotlib inline
import datetime
import importlib
import os
import sys
from pathlib import Path
import numpy as np
import matplotlib.pyplot as plt
import numpy as np
import pandas as pd
import seaborn as sns
from scipy import stats
from sklearn.model_selection import LeaveOneGroupOut, cross_validate
from sklearn.impute import SimpleImputer
from sklearn.dummy import DummyClassifier
from sklearn import linear_model, svm, naive_bayes, neighbors, tree, ensemble
import xgboost as xg
from sklearn.cluster import KMeans
from sklearn.impute import SimpleImputer
from sklearn.model_selection import LeaveOneGroupOut, StratifiedKFold, cross_validate
from IPython.core.interactiveshell import InteractiveShell
InteractiveShell.ast_node_interactivity = "all"
nb_dir = os.path.split(os.getcwd())[0]
if nb_dir not in sys.path:
sys.path.append(nb_dir)
import machine_learning.labels
import machine_learning.model
from machine_learning.classification_models import ClassificationModels
# %% [markdown]
# # RAPIDS models
# %% [markdown]
# %%
# ## Set script's parameters
n_clusters = 4 # Number of clusters (could be regarded as a hyperparameter)
cv_method_str = 'logo' # logo, halflogo, 5kfold # Cross-validation method (could be regarded as a hyperparameter)
n_sl = 1 # Number of largest/smallest accuracies (of particular CV) outputs
N_CLUSTERS = 4 # Number of clusters (could be regarded as a hyperparameter)
CV_METHOD = "logo" # logo, halflogo, 5kfold
# Cross-validation method (could be regarded as a hyperparameter)
N_SL = 1 # Number of largest/smallest accuracies (of particular CV) outputs
# %%
PATH_BASE = Path("E:/STRAWresults/20230415")
SEGMENT_TYPE = "period"
print("SEGMENT_TYPE: " + SEGMENT_TYPE)
SEGMENT_LENGTH = "30_minutes_before"
print("SEGMENT_LENGTH: " + SEGMENT_LENGTH)
TARGET_VARIABLE = "appraisal_stressfulness"
print("TARGET_VARIABLE: " + TARGET_VARIABLE)
if ("appraisal" in TARGET_VARIABLE) and ("stressfulness" in TARGET_VARIABLE):
TARGET_VARIABLE += "_"
TARGET_VARIABLE += SEGMENT_TYPE
PATH_FULL = PATH_BASE / SEGMENT_LENGTH / ("input_" + TARGET_VARIABLE + "_mean.csv")
model_input = pd.read_csv(PATH_FULL)
if SEGMENT_LENGTH == "daily":
DAY_LENGTH = "daily" # or "working"
print(DAY_LENGTH)
model_input = model_input[model_input["local_segment"].str.contains(DAY_LENGTH)]
# %% jupyter={"source_hidden": true}
model_input = pd.read_csv("../data/30min_all_target_inputs/input_JCQ_job_demand_mean.csv")
index_columns = ["local_segment", "local_segment_label", "local_segment_start_datetime", "local_segment_end_datetime"]
index_columns = [
"local_segment",
"local_segment_label",
"local_segment_start_datetime",
"local_segment_end_datetime",
]
clust_col = model_input.set_index(index_columns).var().idxmax() # age is a col with the highest variance
CLUST_COL = "limesurvey_demand_control_ratio_quartile"
print("CLUST_COL: " + CLUST_COL)
model_input.columns[list(model_input.columns).index('age'):-1]
BINS = [-1, 0, 4]
print("BINS: " + str(BINS))
lime_cols = [col for col in model_input if col.startswith('limesurvey')]
lime_cols
lime_col = 'limesurvey_demand_control_ratio_quartile'
clust_col = lime_col
model_input[CLUST_COL].describe()
model_input[clust_col].describe()
# %%
model_input["target"].value_counts()
# %% jupyter={"source_hidden": true}
# Filter-out outlier rows by clust_col
# model_input = model_input[(np.abs(stats.zscore(model_input[clust_col])) < 3)]
# Filter-out outlier rows by clust_col
#model_input = model_input[(np.abs(stats.zscore(model_input[clust_col])) < 3)]
uniq = model_input[[clust_col, 'pid']].drop_duplicates().reset_index(drop=True)
uniq = model_input[[CLUST_COL, "pid"]].drop_duplicates().reset_index(drop=True)
uniq = uniq.dropna()
plt.bar(uniq['pid'], uniq[clust_col])
plt.bar(uniq["pid"], uniq[CLUST_COL])
# %% jupyter={"source_hidden": true}
# Get clusters by cluster col & and merge the clusters to main df
km = KMeans(n_clusters=n_clusters).fit_predict(uniq.set_index('pid'))
km = KMeans(n_clusters=N_CLUSTERS).fit_predict(uniq.set_index("pid"))
np.unique(km, return_counts=True)
uniq['cluster'] = km
uniq
uniq["cluster"] = km
model_input = model_input.merge(uniq[['pid', 'cluster']])
model_input = model_input.merge(uniq[["pid", "cluster"]])
# %%
model_input[["cluster", "target"]].value_counts().sort_index()
# %% jupyter={"source_hidden": true}
model_input.set_index(index_columns, inplace=True)
@ -98,31 +103,56 @@ cm = ClassificationModels()
cmodels = cm.get_cmodels()
# %% jupyter={"source_hidden": true}
for k in range(n_clusters):
for k in range(N_CLUSTERS):
model_input_subset = model_input[model_input["cluster"] == k].copy()
bins = [-10, -1, 1, 10] # bins for z-scored targets
model_input_subset.loc[:, 'target'] = \
pd.cut(model_input_subset.loc[:, 'target'], bins=bins, labels=['low', 'medium', 'high'], right=False) #['low', 'medium', 'high']
model_input_subset['target'].value_counts()
model_input_subset = model_input_subset[model_input_subset['target'] != "medium"]
model_input_subset['target'] = model_input_subset['target'].astype(str).apply(lambda x: 0 if x == "low" else 1)
model_input_subset.loc[:, "target"] = pd.cut(
model_input_subset.loc[:, "target"],
bins=BINS,
labels=["low", "high"],
right=True,
) # ['low', 'medium', 'high']
model_input_subset["target"].value_counts()
# model_input_subset = model_input_subset[model_input_subset["target"] != "medium"]
model_input_subset["target"] = (
model_input_subset["target"].astype(str).apply(lambda x: 0 if x == "low" else 1)
)
model_input_subset['target'].value_counts()
if cv_method_str == 'half_logo':
model_input_subset['pid_index'] = model_input_subset.groupby('pid').cumcount()
model_input_subset['pid_count'] = model_input_subset.groupby('pid')['pid'].transform('count')
print(model_input_subset["target"].value_counts())
model_input_subset["pid_index"] = (model_input_subset['pid_index'] / model_input_subset['pid_count'] + 1).round()
model_input_subset["pid_half"] = model_input_subset["pid"] + "_" + model_input_subset["pid_index"].astype(int).astype(str)
if CV_METHOD == "half_logo":
model_input_subset["pid_index"] = model_input_subset.groupby("pid").cumcount()
model_input_subset["pid_count"] = model_input_subset.groupby("pid")[
"pid"
].transform("count")
data_x, data_y, data_groups = model_input_subset.drop(["target", "pid", "pid_index", "pid_half"], axis=1), model_input_subset["target"], model_input_subset["pid_half"]
model_input_subset["pid_index"] = (
model_input_subset["pid_index"] / model_input_subset["pid_count"] + 1
).round()
model_input_subset["pid_half"] = (
model_input_subset["pid"]
+ "_"
+ model_input_subset["pid_index"].astype(int).astype(str)
)
data_x, data_y, data_groups = (
model_input_subset.drop(["target", "pid", "pid_index", "pid_half"], axis=1),
model_input_subset["target"],
model_input_subset["pid_half"],
)
else:
data_x, data_y, data_groups = model_input_subset.drop(["target", "pid"], axis=1), model_input_subset["target"], model_input_subset["pid"]
data_x, data_y, data_groups = (
model_input_subset.drop(["target", "pid"], axis=1),
model_input_subset["target"],
model_input_subset["pid"],
)
# Treat categorical features
categorical_feature_colnames = ["gender", "startlanguage"]
additional_categorical_features = [col for col in data_x.columns if "mostcommonactivity" in col or "homelabel" in col]
additional_categorical_features = [
col
for col in data_x.columns
if "mostcommonactivity" in col or "homelabel" in col
]
categorical_feature_colnames += additional_categorical_features
categorical_features = data_x[categorical_feature_colnames].copy()
@ -132,7 +162,9 @@ for k in range(n_clusters):
categorical_features = categorical_features.fillna(mode_categorical_features)
# one-hot encoding
categorical_features = categorical_features.apply(lambda col: col.astype("category"))
categorical_features = categorical_features.apply(
lambda col: col.astype("category")
)
if not categorical_features.empty:
categorical_features = pd.get_dummies(categorical_features)
@ -140,8 +172,10 @@ for k in range(n_clusters):
train_x = pd.concat([numerical_features, categorical_features], axis=1)
# Establish cv method
cv_method = StratifiedKFold(n_splits=5, shuffle=True) # Defaults to 5 k-folds in cross_validate method
if cv_method_str == 'logo' or cv_method_str == 'half_logo':
cv_method = StratifiedKFold(
n_splits=5, shuffle=True
) # Defaults to 5 k-folds in cross_validate method
if CV_METHOD == "logo" or CV_METHOD == "half_logo":
cv_method = LeaveOneGroupOut()
cv_method.get_n_splits(
train_x,
@ -149,36 +183,57 @@ for k in range(n_clusters):
groups=data_groups,
)
imputer = SimpleImputer(missing_values=np.nan, strategy='median')
imputer = SimpleImputer(missing_values=np.nan, strategy="median")
for model_title, model in cmodels.items():
classifier = cross_validate(
model['model'],
model["model"],
X=imputer.fit_transform(train_x),
y=data_y,
groups=data_groups,
cv=cv_method,
n_jobs=-1,
error_score='raise',
scoring=('accuracy', 'precision', 'recall', 'f1')
error_score="raise",
scoring=("accuracy", "precision", "recall", "f1"),
)
print("\n-------------------------------------\n")
print("Current cluster:", k, end="\n")
print("Current model:", model_title, end="\n")
print("Acc", np.mean(classifier['test_accuracy']))
print("Precision", np.mean(classifier['test_precision']))
print("Recall", np.mean(classifier['test_recall']))
print("F1", np.mean(classifier['test_f1']))
print(f"Largest {n_sl} ACC:", np.sort(-np.partition(-classifier['test_accuracy'], n_sl)[:n_sl])[::-1])
print(f"Smallest {n_sl} ACC:", np.sort(np.partition(classifier['test_accuracy'], n_sl)[:n_sl]))
cmodels[model_title]['metrics'][0] += np.mean(classifier['test_accuracy'])
cmodels[model_title]['metrics'][1] += np.mean(classifier['test_precision'])
cmodels[model_title]['metrics'][2] += np.mean(classifier['test_recall'])
cmodels[model_title]['metrics'][3] += np.mean(classifier['test_f1'])
print("Acc", np.mean(classifier["test_accuracy"]))
print("Precision", np.mean(classifier["test_precision"]))
print("Recall", np.mean(classifier["test_recall"]))
print("F1", np.mean(classifier["test_f1"]))
print(
f"Largest {N_SL} ACC:",
np.sort(-np.partition(-classifier["test_accuracy"], N_SL)[:N_SL])[::-1],
)
print(
f"Smallest {N_SL} ACC:",
np.sort(np.partition(classifier["test_accuracy"], N_SL)[:N_SL]),
)
cmodels[model_title]["metrics"][0] += np.mean(classifier["test_accuracy"])
cmodels[model_title]["metrics"][1] += np.mean(classifier["test_precision"])
cmodels[model_title]["metrics"][2] += np.mean(classifier["test_recall"])
cmodels[model_title]["metrics"][3] += np.mean(classifier["test_f1"])
# %% jupyter={"source_hidden": true}
# Get overall results
cm.get_total_models_scores(n_clusters=n_clusters)
scores = cm.get_total_models_scores(n_clusters=N_CLUSTERS)
# %%
PATH_OUTPUT = Path("..") / Path("presentation/results")
path_output_full = PATH_OUTPUT / (
TARGET_VARIABLE
+ "_"
+ SEGMENT_LENGTH
+ "_classification_"
+ CV_METHOD
+ str(BINS)
+ "_clust_"
+ CLUST_COL
+ str(N_CLUSTERS)
+ ".csv"
)
scores.to_csv(path_output_full, index=False)

View File

@ -6,7 +6,7 @@
# extension: .py
# format_name: percent
# format_version: '1.3'
# jupytext_version: 1.13.0
# jupytext_version: 1.14.5
# kernelspec:
# display_name: straw2analysis
# language: python
@ -14,92 +14,83 @@
# ---
# %% jupyter={"source_hidden": true}
# %matplotlib inline
import os
import sys
from pathlib import Path
import numpy as np
import matplotlib.pyplot as plt
import numpy as np
import pandas as pd
from scipy import stats
from sklearn.model_selection import train_test_split
from sklearn.impute import SimpleImputer
from sklearn.metrics import accuracy_score, precision_score, recall_score, f1_score
from sklearn.cluster import KMeans
from IPython.core.interactiveshell import InteractiveShell
InteractiveShell.ast_node_interactivity = "all"
nb_dir = os.path.split(os.getcwd())[0]
if nb_dir not in sys.path:
sys.path.append(nb_dir)
from sklearn.impute import SimpleImputer
from sklearn.metrics import accuracy_score, f1_score, precision_score, recall_score
from sklearn.model_selection import train_test_split
from machine_learning.classification_models import ClassificationModels
# %% [markdown]
# # RAPIDS models
# %% [markdown]
# # Useful method
def treat_categorical_features(input_set):
categorical_feature_colnames = ["gender", "startlanguage"]
additional_categorical_features = [col for col in input_set.columns if "mostcommonactivity" in col or "homelabel" in col]
categorical_feature_colnames += additional_categorical_features
categorical_features = input_set[categorical_feature_colnames].copy()
mode_categorical_features = categorical_features.mode().iloc[0]
# fillna with mode
categorical_features = categorical_features.fillna(mode_categorical_features)
# one-hot encoding
categorical_features = categorical_features.apply(lambda col: col.astype("category"))
if not categorical_features.empty:
categorical_features = pd.get_dummies(categorical_features)
numerical_features = input_set.drop(categorical_feature_colnames, axis=1)
return pd.concat([numerical_features, categorical_features], axis=1)
from machine_learning.helper import impute_encode_categorical_features
# %% [markdown]
# ## Set script's parameters
n_clusters = 3 # Number of clusters (could be regarded as a hyperparameter)
n_sl = 3 # Number of largest/smallest accuracies (of particular CV) outputs
#
# %%
n_clusters = 3 # Number of clusters (could be regarded as a hyperparameter)
n_sl = 3 # Number of largest/smallest accuracies (of particular CV) outputs
# %%
PATH_BASE = Path("E:/STRAWresults/20230415")
SEGMENT_TYPE = "period"
print("SEGMENT_TYPE: " + SEGMENT_TYPE)
SEGMENT_LENGTH = "30_minutes_before"
print("SEGMENT_LENGTH: " + SEGMENT_LENGTH)
TARGET_VARIABLE = "appraisal_stressfulness"
print("TARGET_VARIABLE: " + TARGET_VARIABLE)
if ("appraisal" in TARGET_VARIABLE) and ("stressfulness" in TARGET_VARIABLE):
TARGET_VARIABLE += "_"
TARGET_VARIABLE += SEGMENT_TYPE
PATH_FULL = PATH_BASE / SEGMENT_LENGTH / ("input_" + TARGET_VARIABLE + "_mean.csv")
model_input = pd.read_csv(PATH_FULL)
if SEGMENT_LENGTH == "daily":
DAY_LENGTH = "daily" # or "working"
print(DAY_LENGTH)
model_input = model_input[model_input["local_segment"].str.contains(DAY_LENGTH)]
# %% jupyter={"source_hidden": true}
model_input = pd.read_csv("../data/intradaily_30_min_all_targets/input_JCQ_job_demand_mean.csv")
index_columns = ["local_segment", "local_segment_label", "local_segment_start_datetime", "local_segment_end_datetime"]
CLUST_COL = "limesurvey_demand_control_ratio"
print("CLUST_COL: " + CLUST_COL)
clust_col = model_input.set_index(index_columns).var().idxmax() # age is a col with the highest variance
BINS = [-1, 0, 4]
print("BINS: " + str(BINS))
model_input.columns[list(model_input.columns).index('age'):-1]
index_columns = [
"local_segment",
"local_segment_label",
"local_segment_start_datetime",
"local_segment_end_datetime",
]
lime_cols = [col for col in model_input if col.startswith('limesurvey')]
lime_cols
lime_col = 'limesurvey_demand_control_ratio'
clust_col = lime_col
model_input[clust_col].describe()
model_input[CLUST_COL].describe()
# %% jupyter={"source_hidden": true}
# Filter-out outlier rows by clust_col
model_input = model_input[(np.abs(stats.zscore(model_input[CLUST_COL])) < 3)]
# Filter-out outlier rows by clust_col
model_input = model_input[(np.abs(stats.zscore(model_input[clust_col])) < 3)]
uniq = model_input[[clust_col, 'pid']].drop_duplicates().reset_index(drop=True)
plt.bar(uniq['pid'], uniq[clust_col])
uniq = model_input[[CLUST_COL, "pid"]].drop_duplicates().reset_index(drop=True)
plt.bar(uniq["pid"], uniq[CLUST_COL])
# %% jupyter={"source_hidden": true}
# Get clusters by cluster col & and merge the clusters to main df
km = KMeans(n_clusters=n_clusters).fit_predict(uniq.set_index('pid'))
km = KMeans(n_clusters=n_clusters).fit_predict(uniq.set_index("pid"))
np.unique(km, return_counts=True)
uniq['cluster'] = km
uniq
uniq["cluster"] = km
print(uniq)
model_input = model_input.merge(uniq[['pid', 'cluster']])
model_input = model_input.merge(uniq[["pid", "cluster"]])
# %% jupyter={"source_hidden": true}
model_input.set_index(index_columns, inplace=True)
@ -109,50 +100,64 @@ model_input.set_index(index_columns, inplace=True)
cm = ClassificationModels()
cmodels = cm.get_cmodels()
# %%
model_input["target"].value_counts()
# %% jupyter={"source_hidden": true}
for k in range(n_clusters):
model_input_subset = model_input[model_input["cluster"] == k].copy()
# Takes 10th percentile and above 90th percentile as the test set -> the rest for the training set. Only two classes, seperated by z-score of 0.
model_input_subset['numerical_target'] = model_input_subset['target']
bins = [-10, 0, 10] # bins for z-scored targets
model_input_subset.loc[:, 'target'] = \
pd.cut(model_input_subset.loc[:, 'target'], bins=bins, labels=[0, 1], right=True)
p15 = np.percentile(model_input_subset['numerical_target'], 15)
p85 = np.percentile(model_input_subset['numerical_target'], 85)
# Treat categorical features
model_input_subset = treat_categorical_features(model_input_subset)
# Split to train, validate, and test subsets
train_set = model_input_subset[(model_input_subset['numerical_target'] > p15) & (model_input_subset['numerical_target'] < p85)].drop(['numerical_target'], axis=1)
test_set = model_input_subset[(model_input_subset['numerical_target'] <= p15) | (model_input_subset['numerical_target'] >= p85)].drop(['numerical_target'], axis=1)
train_set['target'].value_counts()
test_set['target'].value_counts()
# Takes 10th percentile and above 90th percentile as the test set -> the rest for the training set. Only two classes, seperated by z-score of 0.
# model_input_subset['numerical_target'] = model_input_subset['target']
model_input_subset.loc[:, "target"] = pd.cut(
model_input_subset.loc[:, "target"], bins=BINS, labels=[0, 1], right=True
)
# p15 = np.percentile(model_input_subset['numerical_target'], 15)
# p85 = np.percentile(model_input_subset['numerical_target'], 85)
# Treat categorical features
model_input_subset = impute_encode_categorical_features(model_input_subset)
# Split to train, validate, and test subsets
# train_set = model_input_subset[(model_input_subset['numerical_target'] > p15) & (model_input_subset['numerical_target'] < p85)].drop(['numerical_target'], axis=1)
# test_set = model_input_subset[(model_input_subset['numerical_target'] <= p15) | (model_input_subset['numerical_target'] >= p85)].drop(['numerical_target'], axis=1)
train_set, test_set = train_test_split(
model_input_subset,
test_size=0.3,
stratify=model_input_subset["pid"],
random_state=42,
)
print(train_set["target"].value_counts())
print(test_set["target"].value_counts())
train_x, train_y = train_set.drop(["target", "pid"], axis=1), train_set["target"]
validate_x, test_x, validate_y, test_y = \
train_test_split(test_set.drop(["target", "pid"], axis=1), test_set["target"], test_size=0.50, random_state=42)
validate_x, test_x, validate_y, test_y = train_test_split(
test_set.drop(["target", "pid"], axis=1),
test_set["target"],
test_size=0.50,
random_state=42,
)
# Impute missing values
imputer = SimpleImputer(missing_values=np.nan, strategy='median')
imputer = SimpleImputer(missing_values=np.nan, strategy="median")
train_x = imputer.fit_transform(train_x)
validate_x = imputer.fit_transform(validate_x)
test_x = imputer.fit_transform(test_x)
for model_title, model in cmodels.items():
model['model'].fit(train_x, train_y)
y_pred = model['model'].predict(validate_x)
model["model"].fit(train_x, train_y)
y_pred = model["model"].predict(validate_x)
acc = accuracy_score(validate_y, y_pred)
prec = precision_score(validate_y, y_pred)
rec = recall_score(validate_y, y_pred)
f1 = f1_score(validate_y, y_pred)
print("\n-------------------------------------\n")
print("Current cluster:", k, end="\n")
print("Current model:", model_title, end="\n")
@ -160,12 +165,30 @@ for k in range(n_clusters):
print("Precision", prec)
print("Recall", rec)
print("F1", f1)
cmodels[model_title]['metrics'][0] += acc
cmodels[model_title]['metrics'][1] += prec
cmodels[model_title]['metrics'][2] += rec
cmodels[model_title]['metrics'][3] += f1
cmodels[model_title]["metrics"][0] += acc
cmodels[model_title]["metrics"][1] += prec
cmodels[model_title]["metrics"][2] += rec
cmodels[model_title]["metrics"][3] += f1
# %% jupyter={"source_hidden": true}
# Get overall results
cm.get_total_models_scores(n_clusters=n_clusters)
scores = cm.get_total_models_scores(n_clusters=n_clusters)
# %%
print(scores)
# %%
PATH_OUTPUT = Path("..") / Path("presentation/results")
path_output_full = PATH_OUTPUT / (
TARGET_VARIABLE
+ "_"
+ SEGMENT_LENGTH
+ "_classification"
+ str(BINS)
+ "_CLUST_"
+ CLUST_COL
+ +str(n_clusters)
+ ".csv"
)
scores.to_csv(path_output_full, index=False)

View File

@ -6,445 +6,61 @@
# extension: .py
# format_name: percent
# format_version: '1.3'
# jupytext_version: 1.13.0
# jupytext_version: 1.14.5
# kernelspec:
# display_name: straw2analysis
# language: python
# name: straw2analysis
# ---
# %% jupyter={"source_hidden": true}
# %matplotlib inline
# %%
import os
import sys
import numpy as np
import pandas as pd
import xgboost as xg
from machine_learning.helper import prepare_regression_model_input
from sklearn import gaussian_process, kernel_ridge, linear_model, svm
from sklearn.dummy import DummyRegressor
from sklearn.impute import SimpleImputer
from sklearn.model_selection import LeaveOneGroupOut, cross_validate
# from IPython.core.interactiveshell import InteractiveShell
# InteractiveShell.ast_node_interactivity = "all"
from machine_learning.helper import (
impute_encode_categorical_features,
prepare_cross_validator,
prepare_sklearn_data_format,
run_all_regression_models,
)
nb_dir = os.path.split(os.getcwd())[0]
if nb_dir not in sys.path:
sys.path.append(nb_dir)
# %% jupyter={"source_hidden": true}
# %%
model_input = pd.read_csv(
"../data/intradaily_30_min_all_targets/input_JCQ_job_demand_mean.csv"
)
# %% jupyter={"source_hidden": true}
cv_method = "half_logo" # logo, half_logo, 5kfold
train_x, data_y, data_groups = prepare_regression_model_input(model_input, cv_method)
# %% jupyter={"source_hidden": true}
logo = LeaveOneGroupOut()
logo.get_n_splits(
train_x,
data_y,
groups=data_groups,
)
# Defaults to 5 k folds in cross_validate method
if cv_method != "logo" and cv_method != "half_logo":
logo = None
# %% jupyter={"source_hidden": true}
sum(data_y.isna())
# %% [markdown]
# ### Baseline: Dummy Regression (mean)
dummy_regr = DummyRegressor(strategy="mean")
# %% jupyter={"source_hidden": true}
imputer = SimpleImputer(missing_values=np.nan, strategy="mean")
# %% jupyter={"source_hidden": true}
dummy_regressor = cross_validate(
dummy_regr,
X=imputer.fit_transform(train_x),
y=data_y,
groups=data_groups,
cv=logo,
n_jobs=-1,
scoring=(
"r2",
"neg_mean_squared_error",
"neg_mean_absolute_error",
"neg_root_mean_squared_error",
),
)
print(
"Negative Mean Squared Error",
np.median(dummy_regressor["test_neg_mean_squared_error"]),
)
print(
"Negative Mean Absolute Error",
np.median(dummy_regressor["test_neg_mean_absolute_error"]),
)
print(
"Negative Root Mean Squared Error",
np.median(dummy_regressor["test_neg_root_mean_squared_error"]),
)
print("R2", np.median(dummy_regressor["test_r2"]))
# %% [markdown]
# ### Linear Regression
# %% jupyter={"source_hidden": true}
lin_reg_rapids = linear_model.LinearRegression()
# %% jupyter={"source_hidden": true}
imputer = SimpleImputer(missing_values=np.nan, strategy="mean")
# %% jupyter={"source_hidden": true}
lin_reg_scores = cross_validate(
lin_reg_rapids,
X=imputer.fit_transform(train_x),
y=data_y,
groups=data_groups,
cv=logo,
n_jobs=-1,
scoring=(
"r2",
"neg_mean_squared_error",
"neg_mean_absolute_error",
"neg_root_mean_squared_error",
),
)
print(
"Negative Mean Squared Error",
np.median(lin_reg_scores["test_neg_mean_squared_error"]),
)
print(
"Negative Mean Absolute Error",
np.median(lin_reg_scores["test_neg_mean_absolute_error"]),
)
print(
"Negative Root Mean Squared Error",
np.median(lin_reg_scores["test_neg_root_mean_squared_error"]),
)
print("R2", np.median(lin_reg_scores["test_r2"]))
# %% [markdown]
# ### XGBRegressor Linear Regression
# %% jupyter={"source_hidden": true}
xgb_r = xg.XGBRegressor(objective="reg:squarederror", n_estimators=10)
# %% jupyter={"source_hidden": true}
imputer = SimpleImputer(missing_values=np.nan, strategy="mean")
# %% jupyter={"source_hidden": true}
xgb_reg_scores = cross_validate(
xgb_r,
X=imputer.fit_transform(train_x),
y=data_y,
groups=data_groups,
cv=logo,
n_jobs=-1,
scoring=(
"r2",
"neg_mean_squared_error",
"neg_mean_absolute_error",
"neg_root_mean_squared_error",
),
)
print(
"Negative Mean Squared Error",
np.median(xgb_reg_scores["test_neg_mean_squared_error"]),
)
print(
"Negative Mean Absolute Error",
np.median(xgb_reg_scores["test_neg_mean_absolute_error"]),
)
print(
"Negative Root Mean Squared Error",
np.median(xgb_reg_scores["test_neg_root_mean_squared_error"]),
)
print("R2", np.median(xgb_reg_scores["test_r2"]))
# %% [markdown]
# ### XGBRegressor Pseudo Huber Error Regression
# %% jupyter={"source_hidden": true}
xgb_psuedo_huber_r = xg.XGBRegressor(objective="reg:pseudohubererror", n_estimators=10)
# %% jupyter={"source_hidden": true}
imputer = SimpleImputer(missing_values=np.nan, strategy="mean")
# %% jupyter={"source_hidden": true}
xgb_psuedo_huber_reg_scores = cross_validate(
xgb_psuedo_huber_r,
X=imputer.fit_transform(train_x),
y=data_y,
groups=data_groups,
cv=logo,
n_jobs=-1,
scoring=(
"r2",
"neg_mean_squared_error",
"neg_mean_absolute_error",
"neg_root_mean_squared_error",
),
)
print(
"Negative Mean Squared Error",
np.median(xgb_psuedo_huber_reg_scores["test_neg_mean_squared_error"]),
)
print(
"Negative Mean Absolute Error",
np.median(xgb_psuedo_huber_reg_scores["test_neg_mean_absolute_error"]),
)
print(
"Negative Root Mean Squared Error",
np.median(xgb_psuedo_huber_reg_scores["test_neg_root_mean_squared_error"]),
)
print("R2", np.median(xgb_psuedo_huber_reg_scores["test_r2"]))
# %% [markdown]
# ### Ridge regression
# %% jupyter={"source_hidden": true}
ridge_reg = linear_model.Ridge(alpha=0.5)
# %% tags=[] jupyter={"source_hidden": true}
ridge_reg_scores = cross_validate(
ridge_reg,
X=imputer.fit_transform(train_x),
y=data_y,
groups=data_groups,
cv=logo,
n_jobs=-1,
scoring=(
"r2",
"neg_mean_squared_error",
"neg_mean_absolute_error",
"neg_root_mean_squared_error",
),
)
print(
"Negative Mean Squared Error",
np.median(ridge_reg_scores["test_neg_mean_squared_error"]),
)
print(
"Negative Mean Absolute Error",
np.median(ridge_reg_scores["test_neg_mean_absolute_error"]),
)
print(
"Negative Root Mean Squared Error",
np.median(ridge_reg_scores["test_neg_root_mean_squared_error"]),
)
print("R2", np.median(ridge_reg_scores["test_r2"]))
# %% [markdown]
# ### Lasso
# %% jupyter={"source_hidden": true}
lasso_reg = linear_model.Lasso(alpha=0.1)
# %% jupyter={"source_hidden": true}
lasso_reg_score = cross_validate(
lasso_reg,
X=imputer.fit_transform(train_x),
y=data_y,
groups=data_groups,
cv=logo,
n_jobs=-1,
scoring=(
"r2",
"neg_mean_squared_error",
"neg_mean_absolute_error",
"neg_root_mean_squared_error",
),
)
print(
"Negative Mean Squared Error",
np.median(lasso_reg_score["test_neg_mean_squared_error"]),
)
print(
"Negative Mean Absolute Error",
np.median(lasso_reg_score["test_neg_mean_absolute_error"]),
)
print(
"Negative Root Mean Squared Error",
np.median(lasso_reg_score["test_neg_root_mean_squared_error"]),
)
print("R2", np.median(lasso_reg_score["test_r2"]))
# %% [markdown]
# ### Bayesian Ridge
# %% jupyter={"source_hidden": true}
bayesian_ridge_reg = linear_model.BayesianRidge()
# %% jupyter={"source_hidden": true}
bayesian_ridge_reg_score = cross_validate(
bayesian_ridge_reg,
X=imputer.fit_transform(train_x),
y=data_y,
groups=data_groups,
cv=logo,
n_jobs=-1,
scoring=(
"r2",
"neg_mean_squared_error",
"neg_mean_absolute_error",
"neg_root_mean_squared_error",
),
)
print(
"Negative Mean Squared Error",
np.median(bayesian_ridge_reg_score["test_neg_mean_squared_error"]),
)
print(
"Negative Mean Absolute Error",
np.median(bayesian_ridge_reg_score["test_neg_mean_absolute_error"]),
)
print(
"Negative Root Mean Squared Error",
np.median(bayesian_ridge_reg_score["test_neg_root_mean_squared_error"]),
)
print("R2", np.median(bayesian_ridge_reg_score["test_r2"]))
# %% [markdown]
# ### RANSAC (outlier robust regression)
# %% jupyter={"source_hidden": true}
ransac_reg = linear_model.RANSACRegressor()
# %% jupyter={"source_hidden": true}
ransac_reg_scores = cross_validate(
ransac_reg,
X=imputer.fit_transform(train_x),
y=data_y,
groups=data_groups,
cv=logo,
n_jobs=-1,
scoring=(
"r2",
"neg_mean_squared_error",
"neg_mean_absolute_error",
"neg_root_mean_squared_error",
),
)
print(
"Negative Mean Squared Error",
np.median(ransac_reg_scores["test_neg_mean_squared_error"]),
)
print(
"Negative Mean Absolute Error",
np.median(ransac_reg_scores["test_neg_mean_absolute_error"]),
)
print(
"Negative Root Mean Squared Error",
np.median(ransac_reg_scores["test_neg_root_mean_squared_error"]),
)
print("R2", np.median(ransac_reg_scores["test_r2"]))
# %% [markdown]
# ### Support vector regression
# %% jupyter={"source_hidden": true}
svr = svm.SVR()
# %% jupyter={"source_hidden": true}
svr_scores = cross_validate(
svr,
X=imputer.fit_transform(train_x),
y=data_y,
groups=data_groups,
cv=logo,
n_jobs=-1,
scoring=(
"r2",
"neg_mean_squared_error",
"neg_mean_absolute_error",
"neg_root_mean_squared_error",
),
)
print(
"Negative Mean Squared Error", np.median(svr_scores["test_neg_mean_squared_error"])
)
print(
"Negative Mean Absolute Error",
np.median(svr_scores["test_neg_mean_absolute_error"]),
)
print(
"Negative Root Mean Squared Error",
np.median(svr_scores["test_neg_root_mean_squared_error"]),
)
print("R2", np.median(svr_scores["test_r2"]))
# %% [markdown]
# ### Kernel Ridge regression
# %% jupyter={"source_hidden": true}
kridge = kernel_ridge.KernelRidge()
# %% jupyter={"source_hidden": true}
kridge_scores = cross_validate(
kridge,
X=imputer.fit_transform(train_x),
y=data_y,
groups=data_groups,
cv=logo,
n_jobs=-1,
scoring=(
"r2",
"neg_mean_squared_error",
"neg_mean_absolute_error",
"neg_root_mean_squared_error",
),
)
print(
"Negative Mean Squared Error",
np.median(kridge_scores["test_neg_mean_squared_error"]),
)
print(
"Negative Mean Absolute Error",
np.median(kridge_scores["test_neg_mean_absolute_error"]),
)
print(
"Negative Root Mean Squared Error",
np.median(kridge_scores["test_neg_root_mean_squared_error"]),
)
print("R2", np.median(kridge_scores["test_r2"]))
# %% [markdown]
# ### Gaussian Process Regression
# %% jupyter={"source_hidden": true}
gpr = gaussian_process.GaussianProcessRegressor()
# %% jupyter={"source_hidden": true}
gpr_scores = cross_validate(
gpr,
X=imputer.fit_transform(train_x),
y=data_y,
groups=data_groups,
cv=logo,
n_jobs=-1,
scoring=(
"r2",
"neg_mean_squared_error",
"neg_mean_absolute_error",
"neg_root_mean_squared_error",
),
)
print(
"Negative Mean Squared Error", np.median(gpr_scores["test_neg_mean_squared_error"])
)
print(
"Negative Mean Absolute Error",
np.median(gpr_scores["test_neg_mean_absolute_error"]),
)
print(
"Negative Root Mean Squared Error",
np.median(gpr_scores["test_neg_root_mean_squared_error"]),
)
print("R2", np.median(gpr_scores["test_r2"]))
# %%
model_input = model_input[model_input["local_segment"].str.contains("daily")]
# %%
CV_METHOD = "logo" # logo, half_logo, 5kfold
model_input_encoded = impute_encode_categorical_features(model_input)
# %%
data_x, data_y, data_groups = prepare_sklearn_data_format(
model_input_encoded, CV_METHOD
)
cross_validator = prepare_cross_validator(data_x, data_y, data_groups, CV_METHOD)
# %%
data_y.head()
# %%
data_y.tail()
# %%
data_y.shape
# %%
scores = run_all_regression_models(data_x, data_y, data_groups, cross_validator)
# %%
scores.to_csv(
"../presentation/JCQ_supervisor_support_regression_" + CV_METHOD + ".csv",
index=False,
)

View File

@ -0,0 +1,217 @@
# ---
# jupyter:
# jupytext:
# text_representation:
# extension: .py
# format_name: percent
# format_version: '1.3'
# jupytext_version: 1.14.5
# kernelspec:
# display_name: straw2analysis
# language: python
# name: straw2analysis
# ---
# %%
import pandas as pd
from features.esm_JCQ import DICT_JCQ_DEMAND_CONTROL_REVERSE
# %%
limesurvey_questions = pd.read_csv(
"E:/STRAWbaseline/survey637813+question_text.csv", header=None
).T
# %%
limesurvey_questions
# %%
limesurvey_questions[["code", "text"]] = limesurvey_questions[0].str.split(
r"\.\s", expand=True, n=1
)
# %%
limesurvey_questions
# %%
demand_reverse_lime_rows = (
limesurvey_questions["text"].str.startswith(" [Od mene se ne zahteva,")
| limesurvey_questions["text"].str.startswith(" [Imam dovolj časa, da končam")
| limesurvey_questions["text"].str.startswith(
" [Pri svojem delu se ne srečujem s konfliktnimi"
)
)
control_reverse_lime_rows = limesurvey_questions["text"].str.startswith(
" [Moje delo vključuje veliko ponavljajočega"
) | limesurvey_questions["text"].str.startswith(
" [Pri svojem delu imam zelo malo svobode"
)
# %%
demand_reverse_lime = limesurvey_questions[demand_reverse_lime_rows]
demand_reverse_lime.loc[:, "qid"] = demand_reverse_lime["code"].str.extract(
r"\[(\d+)\]"
)
control_reverse_lime = limesurvey_questions[control_reverse_lime_rows]
control_reverse_lime.loc[:, "qid"] = control_reverse_lime["code"].str.extract(
r"\[(\d+)\]"
)
# %%
limesurvey_questions.loc[89, "text"]
# %%
limesurvey_questions[limesurvey_questions["code"].str.startswith("JobEisen")]
# %%
demand_reverse_lime
# %%
control_reverse_lime
# %%
participant_info = pd.read_csv(
"C:/Users/junos/Documents/FWO-ARRS/Analysis/straw2analysis/rapids/data/raw/p031/participant_baseline_raw.csv",
parse_dates=["date_of_birth"],
)
# %%
participant_info_t = participant_info.T
# %%
rows_baseline = participant_info_t.index
# %%
rows_demand = rows_baseline.str.startswith("JobEisen") & ~rows_baseline.str.endswith(
"Time"
)
# %%
rows_baseline[rows_demand]
# %%
limesurvey_control = (
participant_info_t[rows_demand]
.reset_index()
.rename(columns={"index": "question", 0: "score_original"})
)
# %%
limesurvey_control
# %%
limesurvey_control["qid"] = (
limesurvey_control["question"].str.extract(r"\[(\d+)\]").astype(int)
)
# %%
limesurvey_control["question"].str.extract(r"\[(\d+)\]").astype(int)
# %%
limesurvey_control["score"] = limesurvey_control["score_original"]
# %%
limesurvey_control["qid"][0]
# %%
rows_demand_reverse = limesurvey_control["qid"].isin(
DICT_JCQ_DEMAND_CONTROL_REVERSE.keys()
)
limesurvey_control.loc[rows_demand_reverse, "score"] = (
4 + 1 - limesurvey_control.loc[rows_demand_reverse, "score_original"]
)
# %%
JCQ_DEMAND = "JobEisen"
JCQ_CONTROL = "JobControle"
dict_JCQ_demand_control_reverse = {
JCQ_DEMAND: {
3: " [Od mene se ne zahteva,",
4: " [Imam dovolj časa, da končam",
5: " [Pri svojem delu se ne srečujem s konfliktnimi",
},
JCQ_CONTROL: {
2: " |Moje delo vključuje veliko ponavljajočega",
6: " [Pri svojem delu imam zelo malo svobode",
},
}
# %%
limesurvey_control
# %%
test = pd.DataFrame(
data={"question": "one", "score_original": 3, "score": 3, "qid": 10}, index=[0]
)
# %%
pd.concat([test, limesurvey_control]).reset_index()
# %%
limesurvey_control["score"].sum()
# %%
rows_demand_reverse
# %%
dict_JCQ_demand_control_reverse[JCQ_DEMAND].keys()
# %%
limesurvey_control
# %%
DEMAND_CONTROL_RATIO_MIN = 5 / (9 * 4)
DEMAND_CONTROL_RATIO_MAX = (4 * 5) / 9
JCQ_NORMS = {
"F": {
0: DEMAND_CONTROL_RATIO_MIN,
1: 0.45,
2: 0.52,
3: 0.62,
4: DEMAND_CONTROL_RATIO_MAX,
},
"M": {
0: DEMAND_CONTROL_RATIO_MIN,
1: 0.41,
2: 0.48,
3: 0.56,
4: DEMAND_CONTROL_RATIO_MAX,
},
}
# %%
JCQ_NORMS[participant_info.loc[0, "gender"]][0]
# %%
participant_info_t.index.str.startswith("JobControle")
# %%
columns_baseline = participant_info.columns
# %%
columns_demand = columns_baseline.str.startswith(
"JobControle"
) & ~columns_baseline.str.endswith("Time")
# %%
columns_baseline[columns_demand]
# %%
participant_control = participant_info.loc[:, columns_demand]
# %%
participant_control["id"] = participant_control.index
# %%
participant_control
# %%
pd.wide_to_long(
participant_control,
stubnames="JobControle",
i="id",
j="qid",
sep="[",
suffix="(\\d+)]",
)

View File

@ -20,11 +20,47 @@ ANSWER_DAY_OFF = "DayOff3421"
ANSWER_SET_EVENING = "DayFinishedSetEvening"
MAX_MORNING_LENGTH = 3
# When the participants was not yet at work at the time of the first (morning) EMA,
# When the participant was not yet at work at the time of the first (morning) EMA,
# only three items were answered.
# Two sleep related items and one indicating NOT starting work yet.
# Daytime EMAs are all longer, in fact they always consist of at least 6 items.
QUESTIONNAIRE_IDS = {
"sleep_quality": 1,
"PANAS_positive_affect": 8,
"PANAS_negative_affect": 9,
"JCQ_job_demand": 10,
"JCQ_job_control": 11,
"JCQ_supervisor_support": 12,
"JCQ_coworker_support": 13,
"PFITS_supervisor": 14,
"PFITS_coworkers": 15,
"UWES_vigor": 16,
"UWES_dedication": 17,
"UWES_absorption": 18,
"COPE_active": 19,
"COPE_support": 20,
"COPE_emotions": 21,
"balance_life_work": 22,
"balance_work_life": 23,
"recovery_experience_detachment": 24,
"recovery_experience_relaxation": 25,
"symptoms": 26,
"appraisal_stressfulness_event": 87,
"appraisal_threat": 88,
"appraisal_challenge": 89,
"appraisal_event_time": 90,
"appraisal_event_duration": 91,
"appraisal_event_work_related": 92,
"appraisal_stressfulness_period": 93,
"late_work": 94,
"work_hours": 95,
"left_work": 96,
"activities": 97,
"coffee_breaks": 98,
"at_work_yet": 99,
}
def get_esm_data(usernames: Collection) -> pd.DataFrame:
"""
@ -52,8 +88,10 @@ def get_esm_data(usernames: Collection) -> pd.DataFrame:
def preprocess_esm(df_esm: pd.DataFrame) -> pd.DataFrame:
"""
Convert timestamps and expand JSON column.
Convert timestamps into human-readable datetimes and dates
and expand the JSON column into several Pandas DF columns.
and expand the JSON column into several Pandas DF columns.
Parameters
----------
@ -63,7 +101,8 @@ def preprocess_esm(df_esm: pd.DataFrame) -> pd.DataFrame:
Returns
-------
df_esm_preprocessed: pd.DataFrame
A dataframe with added columns: datetime in Ljubljana timezone and all fields from ESM_JSON column.
A dataframe with added columns: datetime in Ljubljana timezone
and all fields from ESM_JSON column.
"""
df_esm = helper.get_date_from_timestamp(df_esm)
@ -76,31 +115,39 @@ def preprocess_esm(df_esm: pd.DataFrame) -> pd.DataFrame:
def classify_sessions_by_completion(df_esm_preprocessed: pd.DataFrame) -> pd.DataFrame:
"""
For each distinct EMA session, determine how the participant responded to it.
Possible outcomes are: SESSION_STATUS_UNANSWERED, SESSION_STATUS_DAY_FINISHED, and SESSION_STATUS_COMPLETE
Possible outcomes are: SESSION_STATUS_UNANSWERED, SESSION_STATUS_DAY_FINISHED,
and SESSION_STATUS_COMPLETE
This is done in three steps.
First, the esm_status is considered.
If any of the ESMs in a session has a status *other than* "answered", then this session is taken as unfinished.
If any of the ESMs in a session has a status *other than* "answered",
then this session is taken as unfinished.
Second, the sessions which do not represent full questionnaires are identified.
These are sessions where participants only marked they are finished with the day or have not yet started working.
These are sessions where participants only marked they are finished with the day
or have not yet started working.
Third, the sessions with only one item are marked with their trigger.
We never offered questionnaires with single items, so we can be sure these are unfinished.
We never offered questionnaires with single items,
so we can be sure these are unfinished.
Finally, all sessions that remain are marked as completed.
By going through different possibilities in expl_esm_adherence.ipynb, this turned out to be a reasonable option.
By going through different possibilities in expl_esm_adherence.ipynb,
this turned out to be a reasonable option.
Parameters
----------
df_esm_preprocessed: pd.DataFrame
A preprocessed dataframe of esm data, which must include the session ID (esm_session).
A preprocessed dataframe of esm data,
which must include the session ID (esm_session).
Returns
-------
df_session_counts: pd.Dataframe
A dataframe of all sessions (grouped by GROUP_SESSIONS_BY) with their statuses and the number of items.
A dataframe of all sessions (grouped by GROUP_SESSIONS_BY)
with their statuses and the number of items.
"""
sessions_grouped = df_esm_preprocessed.groupby(GROUP_SESSIONS_BY)
@ -155,17 +202,22 @@ def classify_sessions_by_completion(df_esm_preprocessed: pd.DataFrame) -> pd.Dat
def classify_sessions_by_time(df_esm_preprocessed: pd.DataFrame) -> pd.DataFrame:
"""
For each EMA session, determine the time of the first user answer and its time type (morning, workday, or evening.)
Classify EMA sessions into morning, workday, or evening.
For each EMA session, determine the time of the first user answer
and its time type (morning, workday, or evening).
Parameters
----------
df_esm_preprocessed: pd.DataFrame
A preprocessed dataframe of esm data, which must include the session ID (esm_session).
A preprocessed dataframe of esm data,
which must include the session ID (esm_session).
Returns
-------
df_session_time: pd.DataFrame
A dataframe of all sessions (grouped by GROUP_SESSIONS_BY) with their time type and timestamp of first answer.
A dataframe of all sessions (grouped by GROUP_SESSIONS_BY)
with their time type and timestamp of first answer.
"""
df_session_time = (
df_esm_preprocessed.sort_values(["participant_id", "datetime_lj"])
@ -179,13 +231,17 @@ def classify_sessions_by_completion_time(
df_esm_preprocessed: pd.DataFrame,
) -> pd.DataFrame:
"""
The point of this function is to not only classify sessions by using the previously defined functions.
Classify sessions and correct the time type.
The point of this function is to not only classify sessions
by using the previously defined functions.
It also serves to "correct" the time type of some EMA sessions.
A morning questionnaire could seamlessly transition into a daytime questionnaire,
if the participant was already at work.
In this case, the "time" label changed mid-session.
Because of the way classify_sessions_by_time works, this questionnaire was classified as "morning".
Because of the way classify_sessions_by_time works,
this questionnaire was classified as "morning".
But for all intents and purposes, it can be treated as a "daytime" EMA.
The way this scenario is differentiated from a true "morning" questionnaire,
@ -194,13 +250,16 @@ def classify_sessions_by_completion_time(
Parameters
----------
df_esm_preprocessed: pd.DataFrame
A preprocessed dataframe of esm data, which must include the session ID (esm_session).
A preprocessed dataframe of esm data,
which must include the session ID (esm_session).
Returns
-------
df_session_counts_time: pd.DataFrame
A dataframe of all sessions (grouped by GROUP_SESSIONS_BY) with statuses, the number of items,
their time type (with some morning EMAs reclassified) and timestamp of first answer.
A dataframe of all sessions (grouped by GROUP_SESSIONS_BY) with statuses,
the number of items,
their time type (with some morning EMAs reclassified)
and timestamp of first answer.
"""
df_session_counts = classify_sessions_by_completion(df_esm_preprocessed)
@ -219,7 +278,8 @@ def classify_sessions_by_completion_time(
def clean_up_esm(df_esm_preprocessed: pd.DataFrame) -> pd.DataFrame:
"""
This function eliminates invalid ESM responses.
Eliminate invalid ESM responses.
It removes unanswered ESMs and those that indicate end of work and similar.
It also extracts a numeric answer from strings such as "4 - I strongly agree".
@ -256,3 +316,100 @@ def clean_up_esm(df_esm_preprocessed: pd.DataFrame) -> pd.DataFrame:
)
)
return df_esm_clean
def increment_answers(df_esm_clean: pd.DataFrame, increment_by=1):
"""
Increment answers to keep in line with original scoring.
We always used 0 for the lowest value of user answer.
Some scales originally used other scoring, such as starting from 1.
This restores original scoring so that the values are comparable to references.
Parameters
----------
df_esm_clean: pd.DataFrame
A cleaned ESM dataframe, which must also include esm_user_answer_numeric.
increment_by:
A number to add to the user answer.
Returns
-------
df_esm_clean: pd.DataFrame
The same df with addition of a column 'esm_user_answer_numeric'.
"""
try:
df_esm_clean = df_esm_clean.assign(
esm_user_score=lambda x: x.esm_user_answer_numeric + increment_by
)
except AttributeError as e:
print("Please, clean the dataframe first using features.esm.clean_up_esm.")
print(e)
return df_esm_clean
def reassign_question_ids(
df_esm_cleaned: pd.DataFrame, question_ids_content: dict
) -> pd.DataFrame:
"""
Fix question IDs to match their actual content.
Unfortunately, when altering the protocol to adapt to COVID pandemic,
we did not retain original question IDs.
This means that for participants before 2021, they are different
from for the rest of them.
This function searches for question IDs by matching their strings.
Parameters
----------
df_esm_cleaned: pd.DataFrame
A cleaned up dataframe, which must also include esm_user_answer_numeric.
question_ids_content: dict
A dictionary, linking question IDs with their content ("instructions").
Returns
-------
df_esm_fixed: pd.DataFrame
The same dataframe but with fixed question IDs.
"""
df_esm_unique_questions = (
df_esm_cleaned.groupby("question_id")
.esm_instructions.value_counts()
.rename()
.reset_index()
)
# Tabulate all possible answers to each question (group by question ID).
# First, check that we anticipated all esm instructions.
for q_id in question_ids_content.keys():
# Look for all questions ("instructions") occurring in the dataframe.
actual_questions = df_esm_unique_questions.loc[
df_esm_unique_questions["question_id"] == q_id,
"esm_instructions",
]
# These are all answers to a given question (by q_id).
questions_matches = actual_questions.str.startswith(
question_ids_content.get(q_id)
)
# See if they are expected, i.e. included in the dictionary.
if ~actual_questions.all():
print("One of the questions that occur in the data was undefined.")
print("This were the questions found in the data: ")
raise KeyError(actual_questions[~questions_matches])
# In case there is an unexpected answer, raise an exception.
# Next, replace question IDs.
df_esm_fixed = df_esm_cleaned.copy()
df_esm_fixed["question_id"] = df_esm_cleaned["esm_instructions"].apply(
lambda x: next(
(
key
for key, values in question_ids_content.items()
if x.startswith(values)
),
None,
)
)
return df_esm_fixed

View File

@ -0,0 +1,125 @@
COPE_ORIGINAL_MAX = 4
COPE_ORIGINAL_MIN = 1
DICT_COPE_QUESTION_IDS = {
164: (
"I took additional action to try to get rid of the problem",
"Ik deed extra mijn best om er iets aan te doen",
"Vložila sem dodaten napor, da bi rešila problem",
"Vložil sem dodaten napor, da bi rešil problem",
),
165: (
"I concentrated my efforts on doing something about it",
"Ik probeerde de situatie te verbeteren",
"Svoje sile sem usmerila v reševanje nastale situacije",
"Svoje sile sem usmeril v reševanje nastale situacije",
),
166: (
"I did what had to be done, one step at a time",
"Ik deed stap voor stap wat nodig was",
"Naredila sem, kar je bilo potrebno korak za korakom",
"Naredil sem, kar je bilo potrebno korak za korakom",
),
167: (
"I took direct action to get around the problem",
"Ik handelde vlug om het probleem te verhelpen",
"Nekaj sem naredila, da sem zaobšla problem",
"Nekaj sem naredil, da sem zaobšel problem",
),
168: (
"I tried to come up with a strategy about what to do",
"Ik probeerde te verzinnen wat ik er aan kon doen",
"Skušala sem najti ustrezen način za rešitev situacije",
"Skušal sem najti ustrezen način za rešitev situacije",
),
169: (
"I made a plan of action",
"Ik maakte een plan",
"Naredila sem načrt za delovanje",
"Naredil sem načrt za delovanje",
),
170: (
"I thought hard about what steps to take",
"Ik dacht hard na over wat ik moest doen",
"Dobro sem premislila, katere korake moram narediti, da rešim problem",
"Dobro sem premislil, katere korake moram narediti, da rešim problem",
),
171: (
"I thought about how I might best handle the problem",
"lk dacht na over hoe ik het probleem het best kon aanpakken",
"Razmišljala sem, kaj bi bilo najbolje narediti s problemom",
"Razmišljal sem, kaj bi bilo najbolje narediti s problemom",
),
172: (
"I asked people who have had similar experiences what they did",
"Ik vroeg aan mensen met dergelijke ervaringen hoe zij reageerden",
"Vprašala sem posameznike s podobnimi izkušnjami, kaj so storili",
"Vprašal sem posameznike s podobnimi izkušnjami, kaj so storili",
),
173: (
"I tried to get advice from someone about what to do",
"lk vroeg advies aan iemand",
"Pri drugih sem poskušala dobiti nasvet, kaj naj storim",
"Pri drugih sem poskušal dobiti nasvet, kaj naj storim",
),
174: (
"I talked to someone to find out more about the situation",
"Ik sprak met iemand om meer te weten te komen over de situatie",
"Z nekom sem se pogovorila, da bi izvedela še kaj o svojem problemu",
"Z nekom sem se pogovoril, da bi izvedel še kaj o svojem problemu",
),
175: (
"I talked to someone who could do something concrete about the problem",
"Ik sprak met iemand die iets aan het probleem kon doen",
"Pogovorila sem se s kom, ki bi lahko naredil kaj konkretnega",
"Pogovoril sem se s kom, ki bi lahko naredil kaj konkretnega",
),
176: (
"I talked to someone about how I felt",
"Ik sprak met iemand over hoe ik mij voelde",
"Z nekom sem se pogovorila o tem, kako sem se počutila",
"Z nekom sem se pogovoril o tem, kako sem se počutil",
),
177: (
"I tried to get emotional support from friends or relatives",
"Ik zocht steun bij vrienden of familie",
"Skušala sem dobiti čustveno podporo prijateljev ali sorodnikov",
"Skušal sem dobiti čustveno podporo prijateljev ali sorodnikov",
),
178: (
"I discussed my feelings with someone",
"lk besprak mijn gevoelens met iemand",
"O svojih občutkih sem se z nekom pogovorila",
"O svojih občutkih sem se z nekom pogovoril",
),
179: (
"I got sympathy and understanding from someone",
"Ik vroeg medeleven en begrip van iemand",
"Poiskala sem naklonjenost in razumevanje drugih",
"Poiskal sem naklonjenost in razumevanje drugih",
),
180: (
"I got upset and let my emotions out",
"Ik raakte van streek",
"Razburila sem se in to tudi pokazala",
"Razburil sem se in to tudi pokazal",
),
181: (
"I let my feelings out",
"Ik toonde mijn gevoelens",
"Svojim čustvom sem dala prosto pot",
"Svojim čustvom sem dal prosto pot",
),
182: (
"I felt a lot of emotional distress and I found myself expressing",
"lk liet duidelijk blijken hoe ellendig ik mij voelde",
"Doživljala sem veliko stresa in opažala, da sem čustva",
"Doživljal sem veliko stresa in opažal, da sem čustva",
),
183: (
"I got upset, and I was really aware of it",
"Ik merkte dat ik erg van streek was",
"Razburila sem se in razmišljala samo o tem",
"Razburil sem se in razmišljal samo o tem",
),
}

View File

@ -1,9 +1,11 @@
import pandas as pd
from features.esm import increment_answers
JCQ_ORIGINAL_MAX = 4
JCQ_ORIGINAL_MIN = 1
dict_JCQ_demand_control_reverse = {
DICT_JCQ_DEMAND_CONTROL_REVERSE = {
75: (
"I was NOT asked",
"Men legde mij geen overdreven",
@ -40,10 +42,14 @@ def reverse_jcq_demand_control_scoring(
df_esm_jcq_demand_control: pd.DataFrame,
) -> pd.DataFrame:
"""
This function recodes answers in Job content questionnaire by first incrementing them by 1,
to be in line with original (1-4) scoring.
Then, some answers are reversed (i.e. 1 becomes 4 etc.), because the questions are negatively phrased.
These answers are listed in dict_JCQ_demand_control_reverse and identified by their question ID.
Reverse JCQ demand and control answers.
This function recodes answers in Job content questionnaire
by first incrementing them by 1, to be in line with original (1-4) scoring.
Then, some answers are reversed (i.e. 1 becomes 4 etc.),
because the questions are negatively phrased.
These answers are listed in DICT_JCQ_DEMAND_CONTROL_REVERSE
and identified by their question ID.
However, the existing data is checked against literal phrasing of these questions
to protect against wrong numbering of questions (differing question IDs).
@ -55,7 +61,8 @@ def reverse_jcq_demand_control_scoring(
Returns
-------
df_esm_jcq_demand_control: pd.DataFrame
The same dataframe with a column esm_user_score containing answers recoded and reversed.
The same dataframe with a column esm_user_score
containing answers recoded and reversed.
"""
df_esm_jcq_demand_control_unique_answers = (
df_esm_jcq_demand_control.groupby("question_id")
@ -64,7 +71,7 @@ def reverse_jcq_demand_control_scoring(
.reset_index()
)
# Tabulate all possible answers to each question (group by question ID).
for q_id in dict_JCQ_demand_control_reverse.keys():
for q_id in DICT_JCQ_DEMAND_CONTROL_REVERSE.keys():
# Look through all answers that need to be reversed.
possible_answers = df_esm_jcq_demand_control_unique_answers.loc[
df_esm_jcq_demand_control_unique_answers["question_id"] == q_id,
@ -72,7 +79,7 @@ def reverse_jcq_demand_control_scoring(
]
# These are all answers to a given question (by q_id).
answers_matches = possible_answers.str.startswith(
dict_JCQ_demand_control_reverse.get(q_id)
DICT_JCQ_DEMAND_CONTROL_REVERSE.get(q_id)
)
# See if they are expected, i.e. included in the dictionary.
if ~answers_matches.all():
@ -82,18 +89,16 @@ def reverse_jcq_demand_control_scoring(
# In case there is an unexpected answer, raise an exception.
try:
df_esm_jcq_demand_control = df_esm_jcq_demand_control.assign(
esm_user_score=lambda x: x.esm_user_answer_numeric + 1
)
# Increment the original answer by 1
# to keep in line with traditional scoring (JCQ_ORIGINAL_MIN - JCQ_ORIGINAL_MAX).
df_esm_jcq_demand_control = increment_answers(df_esm_jcq_demand_control)
# Increment the original answer by 1 to keep in line
# with traditional scoring (from JCQ_ORIGINAL_MIN to JCQ_ORIGINAL_MAX).
df_esm_jcq_demand_control[
df_esm_jcq_demand_control["question_id"].isin(
dict_JCQ_demand_control_reverse.keys()
DICT_JCQ_DEMAND_CONTROL_REVERSE.keys()
)
] = df_esm_jcq_demand_control[
df_esm_jcq_demand_control["question_id"].isin(
dict_JCQ_demand_control_reverse.keys()
DICT_JCQ_DEMAND_CONTROL_REVERSE.keys()
)
].assign(
esm_user_score=lambda x: JCQ_ORIGINAL_MAX

View File

@ -3,6 +3,9 @@ import pandas as pd
import features.esm
SAM_ORIGINAL_MAX = 5
SAM_ORIGINAL_MIN = 1
QUESTIONNAIRE_ID_SAM = {
"event_stress": 87,
"event_threat": 88,
@ -20,10 +23,107 @@ GROUP_QUESTIONNAIRES_BY = [
"device_id",
"esm_session",
]
# Each questionnaire occurs only once within each esm_session on the same device within the same participant.
# Each questionnaire occurs only once within each esm_session on the same device
# within the same participant.
DICT_SAM_QUESTION_IDS = {
87: (
"Was there a particular event that created tension in you?",
"Was er een bepaalde gebeurtenis die spanning veroorzaakte?",
"Je prišlo do kakega dogodka, ki je v vas ustvaril napetost?",
),
88: (
"Did this event make you feel anxious?",
"Voelde je je angstig door deze gebeurtenis?",
"Ste se zaradi tega dogodka počutili tesnob",
),
89: (
"Will the outcome of this event be negative?",
"Zal de uitkomst van deze gebeurtenis negatief zijn?",
"Bo izid tega dogodka negativen?",
),
90: (
"How threatening was this event?",
"Hoe bedreigend was deze gebeurtenis?",
"Kako grozeč je bil ta dogodek?",
),
91: (
"Is this going to have a negative impact on you?",
"Zal dit een negatieve impact op je hebben?",
"Ali bo to negativno vplivalo na vas?",
),
92: (
"Is this going to have a positive impact on you?",
"Zal dit een positief effect op je hebben?",
"Ali bo to pozitivno vplivalo na vas?",
),
93: (
"How eager are you to tackle this event?",
"Hoe graag wil je deze gebeurtenis aanpakken?",
"Kako zagnani ste bili",
),
94: (
"To what extent can you become a stronger person because of this event?",
"In welke mate kan je een sterkere persoon worden door deze gebeurtenis?",
"V kolikšni meri lahko zaradi tega dogodka postanete močnejša oseba?",
),
95: (
"To what extent are you excited thinking about the outcome of this event?",
"In welke mate ben je enthousiast bij de gedachte",
"V kolikšni meri vas misel na izid tega dogodka navdušuje?",
),
96: (
"At what time did this event occur?",
"Hoe laat vond deze gebeurtenis plaats?",
"Kdaj se je ta dogodek zgodil?",
),
97: (
"How long did this event last?",
"Hoe lang duurde deze gebeurtenis?",
"Kako dolgo je trajal ta dogodek?",
),
98: (
"Was/is this event work-related?",
"Was/is deze gebeurtenis werkgerelateerd?",
"Je (bil) ta dogodek povezan s službo?",
"Je bil ali je ta dogodek povezan s službo?",
),
99: (
"Did this overall period create tension in you?",
"Heeft deze globale periode spanning veroorzaakt?",
"Je to obdobje kot celota v vas ustvarilo napetost?",
"Je to celo obdobje v vas ustvarilo napetost?",
),
100: (
"To what extent do you perceive this overall period as stressful?",
"In welke mate ervaar je deze globale periode als stressvol?",
"V kolikšni meri ste to obdobje dojemali kot stresno?",
"V kolikšni meri ste celo to obdobje dojemali kot stresno?",
),
}
def extract_stressful_events(df_esm: pd.DataFrame) -> pd.DataFrame:
"""
Extract information about stressful events.
Participants were asked: "Was there a particular event that created tension in you?"
Then a subset of questions related to this event followed.
This function goes through the follow-up questions one by one
and preprocesses them, so that it adds new columns to the dataframe.
Parameters
----------
df_esm: pd.DataFrame
A raw dataframe of all ESM data.
Returns
-------
df_esm_events: pd.DataFrame
A cleaned up df of Stress Appraisal Measure items with additional columns.
"""
# 0. Select only questions from Stress Appraisal Measure.
df_esm_preprocessed = features.esm.preprocess_esm(df_esm)
df_esm_sam = df_esm_preprocessed[
@ -78,7 +178,8 @@ def extract_stressful_events(df_esm: pd.DataFrame) -> pd.DataFrame:
def calculate_threat_challenge_means(df_esm_sam_clean: pd.DataFrame) -> pd.DataFrame:
"""
This function calculates challenge and threat (two Stress Appraisal Measure subscales) means,
This function calculates challenge and threat
(two Stress Appraisal Measure subscales) means,
for each ESM session (within participants and devices).
It creates a grouped dataframe with means in two columns.
@ -90,7 +191,8 @@ def calculate_threat_challenge_means(df_esm_sam_clean: pd.DataFrame) -> pd.DataF
Returns
-------
df_esm_event_threat_challenge_mean_wide: pd.DataFrame
A dataframe of unique ESM sessions (by participants and devices) with threat and challenge means.
A dataframe of unique ESM sessions (by participants and devices)
with threat and challenge means.
"""
# Select only threat and challenge assessments for events
df_esm_event_threat_challenge = df_esm_sam_clean[
@ -112,8 +214,8 @@ def calculate_threat_challenge_means(df_esm_sam_clean: pd.DataFrame) -> pd.DataF
aggfunc="mean",
)
# Drop unnecessary column values.
df_esm_event_threat_challenge_mean_wide.columns = df_esm_event_threat_challenge_mean_wide.columns.get_level_values(
1
df_esm_event_threat_challenge_mean_wide.columns = (
df_esm_event_threat_challenge_mean_wide.columns.get_level_values(1)
)
df_esm_event_threat_challenge_mean_wide.columns.name = None
df_esm_event_threat_challenge_mean_wide.rename(
@ -189,10 +291,12 @@ def detect_event_work_related(df_esm_sam_clean: pd.DataFrame) -> pd.DataFrame:
def convert_event_time(df_esm_sam_clean: pd.DataFrame) -> pd.DataFrame:
"""
This function only serves to convert the string datetime answer into a real datetime type.
Errors during this conversion are coerced, meaning that non-datetime answers are assigned Not a Time (NaT).
NOTE: Since the only available non-datetime answer to this question was "0 - I do not remember",
the NaTs can be interpreted to mean this.
This function only serves to convert the string datetime answer
into a real datetime type.
Errors during this conversion are coerced, meaning that non-datetime answers
are assigned Not a Time (NaT).
NOTE: Since the only available non-datetime answer to this question was
"0 - I do not remember", the NaTs can be interpreted to mean this.
Parameters
----------
@ -208,9 +312,13 @@ def convert_event_time(df_esm_sam_clean: pd.DataFrame) -> pd.DataFrame:
df_esm_sam_clean["questionnaire_id"] == QUESTIONNAIRE_ID_SAM.get("event_time")
].assign(
event_time=lambda x: pd.to_datetime(
x.esm_user_answer, errors="coerce", infer_datetime_format=True, exact=True
x.esm_user_answer,
errors="coerce",
format="%Y-%m-%d %H:%M:%S %z",
exact=True,
)
)
# Example answer: 2020-09-29 00:05:00 +0200
return df_esm_event_time
@ -241,9 +349,12 @@ def extract_event_duration(df_esm_sam_clean: pd.DataFrame) -> pd.DataFrame:
== QUESTIONNAIRE_ID_SAM.get("event_duration")
].assign(
event_duration=lambda x: pd.to_datetime(
x.esm_user_answer.str.slice(start=0, stop=-6), errors="coerce"
x.esm_user_answer.str.slice(start=0, stop=-6),
errors="coerce",
format="%Y-%m-%d %H:%M:%S",
).dt.time
)
# Example answer: 2020-09-29 00:05:00 +0200
# TODO Explore the values recorded in event_duration and possibly fix mistakes.
# For example, participants reported setting 23:50:00 instead of 00:50:00.
@ -251,7 +362,7 @@ def extract_event_duration(df_esm_sam_clean: pd.DataFrame) -> pd.DataFrame:
# we can determine whether:
# - this event is still going on ("1 - It is still going on")
# - the participant couldn't remember it's duration ("0 - I do not remember")
# Generally, these answers were converted to esm_user_answer_numeric in clean_up_esm,
# Generally, these answers were converted to esm_user_answer_numeric in clean_up_esm
# but only the numeric types of questions and answers.
# Since this was of "datetime" type, convert these specific answers here again.
df_esm_event_duration["event_duration_info"] = np.nan
@ -264,4 +375,5 @@ def extract_event_duration(df_esm_sam_clean: pd.DataFrame) -> pd.DataFrame:
return df_esm_event_duration
# TODO: How many questions about the stressfulness of the period were asked and how does this relate to events?
# TODO: How many questions about the stressfulness of the period were asked
# and how does this relate to events?

View File

@ -1,71 +1,123 @@
from sklearn.dummy import DummyClassifier
from sklearn import linear_model, svm, naive_bayes, neighbors, tree, ensemble
import pandas as pd
import xgboost as xg
from lightgbm import LGBMClassifier
import xgboost as xg
from sklearn import ensemble, linear_model, naive_bayes, neighbors, svm, tree
from sklearn.dummy import DummyClassifier
class ClassificationModels():
class ClassificationModels:
def __init__(self):
self.cmodels = self.init_classification_models()
def get_cmodels(self):
return self.cmodels
def init_classification_models(self):
cmodels = {
'dummy_classifier': {
'model': DummyClassifier(strategy="most_frequent"),
'metrics': [0, 0, 0, 0]
"dummy_classifier": {
"model": DummyClassifier(strategy="most_frequent"),
"metrics": [0, 0, 0, 0],
},
'logistic_regression': {
'model': linear_model.LogisticRegression(max_iter=1000),
'metrics': [0, 0, 0, 0]
"logistic_regression": {
"model": linear_model.LogisticRegression(max_iter=1000),
"metrics": [0, 0, 0, 0],
},
'support_vector_machine': {
'model': svm.SVC(),
'metrics': [0, 0, 0, 0]
"support_vector_machine": {"model": svm.SVC(), "metrics": [0, 0, 0, 0]},
"gaussian_naive_bayes": {
"model": naive_bayes.GaussianNB(),
"metrics": [0, 0, 0, 0],
},
'gaussian_naive_bayes': {
'model': naive_bayes.GaussianNB(),
'metrics': [0, 0, 0, 0]
"stochastic_gradient_descent_classifier": {
"model": linear_model.SGDClassifier(),
"metrics": [0, 0, 0, 0],
},
'stochastic_gradient_descent_classifier': {
'model': linear_model.SGDClassifier(),
'metrics': [0, 0, 0, 0]
"knn": {"model": neighbors.KNeighborsClassifier(), "metrics": [0, 0, 0, 0]},
"decision_tree": {
"model": tree.DecisionTreeClassifier(),
"metrics": [0, 0, 0, 0],
},
'knn': {
'model': neighbors.KNeighborsClassifier(),
'metrics': [0, 0, 0, 0]
"random_forest_classifier": {
"model": ensemble.RandomForestClassifier(),
"metrics": [0, 0, 0, 0],
},
'decision_tree': {
'model': tree.DecisionTreeClassifier(),
'metrics': [0, 0, 0, 0]
"gradient_boosting_classifier": {
"model": ensemble.GradientBoostingClassifier(),
"metrics": [0, 0, 0, 0],
},
'random_forest_classifier': {
'model': ensemble.RandomForestClassifier(),
'metrics': [0, 0, 0, 0]
"lgbm_classifier": {"model": LGBMClassifier(), "metrics": [0, 0, 0, 0]},
"XGBoost_classifier": {
"model": xg.sklearn.XGBClassifier(),
"metrics": [0, 0, 0, 0],
},
'gradient_boosting_classifier': {
'model': ensemble.GradientBoostingClassifier(),
'metrics': [0, 0, 0, 0]
},
'lgbm_classifier': {
'model': LGBMClassifier(),
'metrics': [0, 0, 0, 0]
},
'XGBoost_classifier': {
'model': xg.sklearn.XGBClassifier(),
'metrics': [0, 0, 0, 0]
}
}
return cmodels
def get_total_models_scores(self, n_clusters=1):
scores = pd.DataFrame(columns=["method", "metric", "mean"])
for model_title, model in self.cmodels.items():
scores_df = pd.DataFrame(columns=["method", "metric", "mean"])
print("\n************************************\n")
print("Current model:", model_title, end="\n")
print("Acc:", model['metrics'][0]/n_clusters)
print("Precision:", model['metrics'][1]/n_clusters)
print("Recall:", model['metrics'][2]/n_clusters)
print("F1:", model['metrics'][3]/n_clusters)
print("Acc:", model["metrics"][0] / n_clusters)
scores_df = pd.concat(
[
scores_df,
pd.DataFrame(
{
"method": model_title,
"metric": "test_accuracy",
"mean": model["metrics"][0] / n_clusters,
},
index=[0],
),
],
ignore_index=True,
)
print("Precision:", model["metrics"][1] / n_clusters)
scores_df = pd.concat(
[
scores_df,
pd.DataFrame(
{
"method": model_title,
"metric": "test_precision",
"mean": model["metrics"][1] / n_clusters,
},
index=[0],
),
],
ignore_index=True,
)
print("Recall:", model["metrics"][2] / n_clusters)
scores_df = pd.concat(
[
scores_df,
pd.DataFrame(
{
"method": model_title,
"metric": "test_recall",
"mean": model["metrics"][2] / n_clusters,
},
index=[0],
),
],
ignore_index=True,
)
print("F1:", model["metrics"][3] / n_clusters)
scores_df = pd.concat(
[
scores_df,
pd.DataFrame(
{
"method": model_title,
"metric": "test_f1",
"mean": model["metrics"][3] / n_clusters,
},
index=[0],
),
],
ignore_index=True,
)
scores = pd.concat([scores, scores_df])
return scores

View File

@ -49,8 +49,8 @@ class CrossValidation():
data_X, data_y, data_groups = data.drop(["target", "pid", "pid_index", "pid_half"], axis=1), data["target"], data["pid_half"]
elif self.cv_method == "5kfold":
data_X, data_y, data_groups = data.drop(["target", "pid"], axis=1), data["target"], data["pid"]
elif self.cv_method == "Stratified5kfold":
data_X, data_y, data_groups = data.drop(["target", "pid"], axis=1), data["target"], None
self.X, self.y, self.groups = data_X, data_y, data_groups
@ -71,7 +71,7 @@ class CrossValidation():
if self.cv_method in ["logo", "half_logo"]:
self.cv = LeaveOneGroupOut()
elif self.cv_method == "5kfold":
elif self.cv_method == "Stratified5kfold":
self.cv = StratifiedKFold(n_splits=5, shuffle=True)
@ -118,4 +118,11 @@ class CrossValidation():
"""
return self.X.iloc[split[0]], self.y.iloc[split[0]], self.X.iloc[split[1]], self.y.iloc[split[1]]
def get_groups_sets(self, split):
if self.groups is None:
return None, None
else:
return self.groups.iloc[split[0]], self.groups.iloc[split[1]]

View File

@ -1,11 +1,13 @@
import os
import sys
import warnings
import numpy as np
import matplotlib.pyplot as plt
import pandas as pd
from sklearn.feature_selection import SequentialFeatureSelector
from sklearn.feature_selection import SelectKBest, f_classif, mutual_info_classif, f_regression
from sklearn.model_selection import cross_validate, StratifiedKFold, GroupKFold
from sklearn.naive_bayes import GaussianNB
from sklearn.linear_model import Lasso
@ -21,13 +23,15 @@ from sklearn.linear_model import Lasso
class FeatureSelection:
def __init__(self, X_train, X_test, y_train, y_test): # TODO: what about leave-one-subject-out CV?
pass # TODO....
def __init__(self, X, y, groups):
self.X = X
self.y = y
self.groups = groups
def select_best_feature(df, features, method="remove", ml_type="classification", metric="recall", stored_features=[]):
def select_best_feature(self, features, method="remove", ml_category="classification", ml_subcategory="bin", metric="recall", stored_features=[]):
"""The method selects the best feature by testing the prediction on the feature set with or without the current feature.
The "remove" method removes a particular feature and predicts on the test set without it. The "add" method adds a particulat
The "remove" method removes a particular feature and predicts on the test set without it. The "add" method adds a particular
feature to the previously established feature set (stored_features). The best feature is selected dependent on the metric
specified as a parameter.
@ -35,7 +39,11 @@ class FeatureSelection:
df (DataFrame): Input data on which the predictions will be made.
features (list): List of features to select the best/worst from
method (str, optional): remove or add features. Defaults to "remove".
ml_type (str, optional): Either classification or regression ml problem controls the ML algorithm and metric. Defaults to "classification".
ml_category (str, optional): Either classification or regression ml problem controls the ML algorithm and metric.
Defaults to "classification".
ml_subcategory (str, optional): In case of classification '_bin' for binary classification
and 'multi' for multiclass classification. For regression an empty string '' is sufficient.
Defaults to "bin".
metric (str, optional): Selected metric with which the best/worst feature will be determined. Defaults to "recall".
stored_features (list, optional): In case if method is 'add', stored features refer to the features that had been previously added. Defaults to [].
@ -49,173 +57,189 @@ class FeatureSelection:
best_feature = None
if ml_type == "classification" and metric not in ['accuracy', 'precision', 'recall', 'f1']:
raise ValueError("Classification metric not recognized. Please choose 'accuracy', 'precision', 'recall' and/or 'f1'")
elif ml_type == "regression" and metric not in ['r2']:
# Validacije tipov ML in specificiranimi metrikami
if ml_category == "classification":
if ml_subcategory == "bin" and metric not in ['accuracy', 'precision', 'recall', 'f1']:
raise ValueError("Classification metric not recognized. Please choose 'accuracy', 'precision', 'recall' and/or 'f1'")
elif ml_subcategory == "multi":
ml_subcategory_error = False
if metric != "accuracy" and "_" in metric:
metric_s, metric_t = metric.split("_")
if metric_s not in ['accuracy', 'precision', 'recall', 'f1'] or metric_t not in ['micro', 'macro', 'weighted']:
ml_subcategory_error = True
else:
ml_subcategory_error = True
if ml_subcategory_error:
raise ValueError(""""Classification metric for multi-class classification must be specified precisely.
Available metric are: 'accuracy', 'precision', 'recall' and 'f1'.
Only accuracy must be specified as 'accuracy'.
For others please add appropriate suffixes: '_macro', '_micro', or '_weighted', e.g., 'f1_macro'""")
elif ml_category == "regression" and metric not in ['r2']:
raise ValueError("Regression metric not recognized. Please choose 'r2'")
for feat in features:
if method == "remove":
pred_features = [col for col in df.columns if feat != col] # All but feat
pred_features = [col for col in self.X.columns if feat != col] # All but feat
elif method == "add":
pred_features = [feat] + stored_features # Feat with stored features
X, y = df.drop(columns=['target', 'pid'])[pred_features], df['target']
X = self.X[pred_features].copy()
if ml_type == "classification":
if self.groups is not None:
cv = GroupKFold(n_splits=5)
else:
cv = StratifiedKFold(n_splits=5, shuffle=True)
# See link about scoring for multiclassfication
# http://iamirmasoud.com/2022/06/19/understanding-micro-macro-and-weighted-averages-for-scikit-learn-metrics-in-multi-class-classification-with-example/
if ml_category == "classification":
nb = GaussianNB()
model_cv = cross_validate(
nb,
X=X,
y=y,
cv=StratifiedKFold(n_splits=5, shuffle=True),
y=self.y,
cv=cv,
groups=self.groups,
n_jobs=-1,
scoring=('accuracy', 'precision', 'recall', 'f1')
scoring=(metric)
)
with warnings.catch_warnings():
warnings.filterwarnings("ignore", message="Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.")
if metric == "accuracy":
acc = np.mean(model_cv['test_accuracy'])
acc_std = np.std(model_cv['test_accuracy'])
if not best_feature or (acc > best_metric_score):
best_feature = feat
best_metric_score = acc
best_metric_score_std = acc_std
elif metric == "precision":
prec = np.mean(model_cv['test_precision'])
prec_std = np.std(model_cv['test_precision'])
if not best_feature or (prec > best_metric_score):
best_feature = feat
best_metric_score = prec
best_metric_score_std = prec_std
elif metric == "recall":
rec = np.mean(model_cv['test_recall'])
rec_std = np.std(model_cv['test_recall'])
if not best_feature or (rec > best_metric_score):
best_feature = feat
best_metric_score = rec
best_metric_score_std = rec_std
else:
f1 = np.mean(model_cv['test_f1'])
f1_std = np.std(model_cv['test_f1'])
if not best_feature or (f1 > best_metric_score):
best_feature = feat
best_metric_score = f1
best_metric_score_std = f1_std
elif ml_type == "regression":
elif ml_category == "regression":
lass = Lasso()
model_cv = cross_validate(
lass,
X=X,
y=y,
cv=StratifiedKFold(n_splits=5, shuffle=True),
cv=cv,
groups=self.groups,
n_jobs=-1,
scoring=('r2')
)
if metric == "r2":
r2 = np.mean(model_cv['test_r2'])
r2_std = np.std(model_cv['test_r2'])
if not best_feature or (r2 > best_metric_score):
best_feature = feat
best_metric_score = r2
best_metric_score_std = r2_std
else:
raise ValueError("ML type not yet implemented!")
# Section of metrics' scores comparison.
with warnings.catch_warnings():
warnings.filterwarnings("ignore", message="Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.")
metric_score = np.nanmean(model_cv["test_score"])
metric_score_std = np.nanstd(model_cv["test_score"])
if not best_feature or (metric_score > best_metric_score):
best_feature = feat
best_metric_score = metric_score
best_metric_score_std = metric_score_std
return best_feature, best_metric_score, best_metric_score_std
def select_features(df, n_min=20, n_max=50, method="remove", n_not_improve=10):
def select_features(self, n_min=20, n_max=50, k=100, method="remove", ml_type="classification_bin", metric="recall", n_tolerance=10):
"""This method selects a set of features and returns them as a list. It returns number of features
determined in the interval of [n_min, n_max].
n_features = df.shape[1] - 2 # -2 beacause pid and target are not considered
if n_max > n_features:
n_max = n_features
The method consists of two steps:
(1) The method uses sklearn kBest method which selects k best features dependent on the ml_type parameter.
(2) The sequential features removal procedure is executed. Using the remaing features from (1).
The best score is detected using a removal procedure. The procedure sequentially removes the features
that attribute the least to the choosen evaluation metric. If in this sequence the score ML score is
improved the next feature is remove otherwise there is a tolerance criteria (n_tolerance)
with which the next n removed features are inspected whether currently best score is improved.
Args:
n_min (int, optional): Minimal amount of features returned.
n_max (int, optional): Maximal amount of features returned.
k (int, optional): Determines the k in the k-best features method.
If None, SelectKBest feature selection does not execute.
ml_type(str, optional): Type of ML problem. Currently implemented options:
'classification_bin', 'classification_multi', and 'regression_'
method (str, optional): "remove" or "add" features. Defaults to "remove".
n_tolerance (int, optional): If the best score is not improved in n that is specified by this parameter
the method returns index of feature with current best score as a tipping point feature.
Returns:
list: list of selected features
"""
if k is not None and k <= n_max:
raise ValueError("The k parameter needs to be greater than the n_max parameter.")
# Select k-best feature dependent on the type of ML task
ml_category, ml_subcategory = ml_type.split("_")
if k is not None:
if ml_category == "classification":
if ml_subcategory== "bin":
selector = SelectKBest(mutual_info_classif, k=k)
elif ml_subcategory== "multi":
selector = SelectKBest(f_classif, k=k)
else:
raise ValueError("Unknown ML type: cannot recognize ML classification subtype.")
elif ml_category == "regression":
selector = SelectKBest(f_regression, k=k)
else:
raise ValueError("Unknown ML type: cannot recognize ML type. Must be either classification or regression.")
selector.fit(self.X, self.y)
cols_idxs = selector.get_support(indices=True)
self.X = self.X.iloc[:,cols_idxs]
print("All columns (after SelectKBest method):")
print(self.X.columns)
# Sequential feature addition / removal
n_features = self.X.shape[1]
if n_max >= n_features:
n_max = n_features-1 # The algorithm removes at least one feature
if n_min > n_features:
raise ValueError("The number of features in the dataframe must be at least as n_min-1 parameter.")
raise ValueError("The number of remaining features in the dataframe must be at least as n_min+1 parameter.")
if n_max < n_min:
raise ValueError("n_max parameter needs to be greater than or equal to n_min parameter.")
features = df.columns.tolist()
features.remove("pid")
features.remove("target")
features = self.X.columns.tolist()
feature_importance = []
if method == "remove":
best_score = 0
best_feature_indx = None
i_worse = 0
for i in reversed(range(n_features)):
if i+1 == n_min:
break
best_feature, best_metric_score, best_metric_score_std = \
self.select_best_feature(df, features, method=method, ml_type="classification", metric="recall")
feature_importance.append(tuple(i+1, best_feature, best_metric_score, best_metric_score_std))
self.select_best_feature(features, method=method, ml_category=ml_category, ml_subcategory=ml_subcategory, metric=metric)
feature_importance.append((i+1, best_feature, best_metric_score, best_metric_score_std))
features.remove(best_feature)
print("Features left:", i)
if i <= n_max:
if best_metric_score >= best_score:
best_score = best_metric_score
best_feature_indx = i+1
i_worse = 0
else:
i_worse += 1
if i_worse == n_tolerance:
break
feature_importance_df = pd.DataFrame(feature_importance, columns=['i', 'name', 'metric', 'metric_sd'])
# Selekcijski kriterij značilk v rangu max-min
# Npr. izbira najboljšega score-a v tem rangu. Ali pa dokler se v tem rangu score zvišuje za 0.0X, ko se ne izberi tisti set značilk.
# Set značilk se bo izbral od i=1 do i=index_izbrane_značilke
# "Tipping point" značilka mora biti v rangu max-min
selection_area = feature_importance_df[(feature_importance_df["i"] >= n_min+1) & (feature_importance_df["i"] <= n_max)]
selection_area.set_index(["i", "name"], inplace=True)
diffrences = selection_area.diff()
diffrences.dropna(how='any', inplace=True)
# Morda tudi komulativna sumacija? Kjer se preprosto index z najvišjo vrednostjo
cumulative_sumation = diffrences.cumsum()
tipping_feature_indx_1 = cumulative_sumation.idxmax()["metric"]
# Zelo konzervativna metoda, ki ob prvem neizboljšanjem rezultata preneha z iskanjem boljše alternative
tipping_feature_indx_2 = None
for indx, row in diffrences.iterrows():
if row["metric"] > 0:
tipping_feature_indx_2 = indx
else:
break
# Metoda, ki pusti n_not_improve značilkam, da premagajo dosedajno najboljši score
tipping_feature_indx_3 = None
cum_sum_score = 0
i_worse = 0
# TODO: morda bi bilo smisleno združiti diff, cumsum in scores stolpce ...
for indx, row in selection_area.iterrows():
if row["metric"] > 0:
tipping_feature_indx_3 = indx
cum_sum_score += row["metric"]
i_worse = 0
else:
i_worse += 1
if i_worse == n_not_improve:
break
print(feature_importance_df)
print("best_feature_indx", best_feature_indx)
print("best_score", best_score)
features_to_remove = feature_importance_df[feature_importance_df["i"] >= best_feature_indx]["name"].values.tolist()
selected_features = [feat for feat in self.X.columns.tolist() if feat not in features_to_remove]
return selected_features
def make_predictions_with_features(df, groups_substrings, include_group=True, with_cols=[], print_flag=False):
pass
def vizualize_feature_selection_process():
pass
def execute_feature_selection_step():
pass
else:
raise ValueError("Method type not recognized: only the 'remove' method is currently implemented.")

View File

@ -11,7 +11,13 @@ from sklearn import (
svm,
)
from sklearn.dummy import DummyClassifier, DummyRegressor
from sklearn.model_selection import LeaveOneGroupOut, cross_validate
from sklearn.metrics import confusion_matrix
from sklearn.model_selection import (
BaseCrossValidator,
LeaveOneGroupOut,
StratifiedKFold,
cross_validate,
)
from xgboost import XGBClassifier, XGBRegressor
@ -73,7 +79,40 @@ def insert_row(df, row):
return pd.concat([df, pd.DataFrame([row], columns=df.columns)], ignore_index=True)
def prepare_regression_model_input(model_input, cv_method="logo"):
def impute_encode_categorical_features(model_input: pd.DataFrame) -> pd.DataFrame:
categorical_feature_col_names = [
"gender",
"startlanguage",
"limesurvey_demand_control_ratio_quartile",
]
additional_categorical_features = [
col
for col in model_input.columns
if "mostcommonactivity" in col or "homelabel" in col
]
categorical_feature_col_names += additional_categorical_features
categorical_features = model_input[categorical_feature_col_names].copy()
mode_categorical_features = categorical_features.mode().iloc[0]
# fillna with mode
categorical_features = categorical_features.fillna(mode_categorical_features)
# one-hot encoding
categorical_features = categorical_features.apply(
lambda col: col.astype("category")
)
if not categorical_features.empty:
categorical_features = pd.get_dummies(categorical_features)
numerical_features = model_input.drop(categorical_feature_col_names, axis=1)
model_input = pd.concat([numerical_features, categorical_features], axis=1)
return model_input
def prepare_sklearn_data_format(
model_input: pd.DataFrame, cv_method: str = "logo"
) -> tuple:
index_columns = [
"local_segment",
"local_segment_label",
@ -82,13 +121,7 @@ def prepare_regression_model_input(model_input, cv_method="logo"):
]
model_input.set_index(index_columns, inplace=True)
if cv_method == "logo":
data_x, data_y, data_groups = (
model_input.drop(["target", "pid"], axis=1),
model_input["target"],
model_input["pid"],
)
else:
if cv_method == "half_logo":
model_input["pid_index"] = model_input.groupby("pid").cumcount()
model_input["pid_count"] = model_input.groupby("pid")["pid"].transform("count")
@ -104,52 +137,53 @@ def prepare_regression_model_input(model_input, cv_method="logo"):
model_input["target"],
model_input["pid_half"],
)
else:
data_x, data_y, data_groups = (
model_input.drop(["target", "pid"], axis=1),
model_input["target"],
model_input["pid"],
)
return data_x, data_y, data_groups
categorical_feature_colnames = [
"gender",
"startlanguage",
"limesurvey_demand_control_ratio_quartile",
]
additional_categorical_features = [
col
for col in data_x.columns
if "mostcommonactivity" in col or "homelabel" in col
]
categorical_feature_colnames += additional_categorical_features
categorical_features = data_x[categorical_feature_colnames].copy()
def prepare_cross_validator(
data_x: pd.DataFrame,
data_y: pd.DataFrame,
data_groups: pd.DataFrame,
cv_method: str = "logo",
) -> BaseCrossValidator:
if cv_method == "logo" or cv_method == "half_logo":
cv = LeaveOneGroupOut()
cv.get_n_splits(
data_x,
data_y,
groups=data_groups,
)
else:
cv = StratifiedKFold(n_splits=5, shuffle=True)
return cv
mode_categorical_features = categorical_features.mode().iloc[0]
# fillna with mode
categorical_features = categorical_features.fillna(mode_categorical_features)
# one-hot encoding
categorical_features = categorical_features.apply(
lambda col: col.astype("category")
def aggregate_and_transpose(df: pd.DataFrame, statistics=None) -> pd.DataFrame:
if statistics is None:
statistics = ["max", "mean"]
return (
df.agg(statistics)
.transpose()
.reset_index()
.rename(columns={"index": "test_metric"})
)
if not categorical_features.empty:
categorical_features = pd.get_dummies(categorical_features)
numerical_features = data_x.drop(categorical_feature_colnames, axis=1)
train_x = pd.concat([numerical_features, categorical_features], axis=1)
return train_x, data_y, data_groups
def run_all_regression_models(input_csv):
# Prepare data
data_x, data_y, data_groups = prepare_regression_model_input(input_csv)
# Prepare cross validation
logo = LeaveOneGroupOut()
logo.get_n_splits(
data_x,
data_y,
groups=data_groups,
)
def run_all_regression_models(
data_x: pd.DataFrame,
data_y: pd.DataFrame,
data_groups: pd.DataFrame,
cross_validator: BaseCrossValidator,
) -> pd.DataFrame:
metrics = ["r2", "neg_mean_absolute_error", "neg_root_mean_squared_error"]
test_metrics = ["test_" + metric for metric in metrics]
scores = pd.DataFrame(columns=["method", "max", "nanmedian"])
scores = pd.DataFrame(columns=["method", "test_metric", "max", "nanmedian"])
# Validate models
dummy_regr = DummyRegressor(strategy="mean")
@ -158,7 +192,7 @@ def run_all_regression_models(input_csv):
X=data_x,
y=data_y,
groups=data_groups,
cv=logo,
cv=cross_validator,
n_jobs=-1,
scoring=metrics,
)
@ -166,17 +200,19 @@ def run_all_regression_models(input_csv):
print("R^2: ", np.nanmedian(dummy_regr_scores["test_r2"]))
scores_df = pd.DataFrame(dummy_regr_scores)[test_metrics]
scores_df = scores_df.agg(["max", np.nanmedian]).transpose()
scores_df = aggregate_and_transpose(scores_df, statistics=["max", np.nanmedian])
scores_df["method"] = "dummy"
scores = pd.concat([scores, scores_df])
del dummy_regr
del dummy_regr_scores
lin_reg_rapids = linear_model.LinearRegression()
lin_reg = linear_model.LinearRegression()
lin_reg_scores = cross_validate(
lin_reg_rapids,
lin_reg,
X=data_x,
y=data_y,
groups=data_groups,
cv=logo,
cv=cross_validator,
n_jobs=-1,
scoring=metrics,
)
@ -184,9 +220,11 @@ def run_all_regression_models(input_csv):
print("R^2: ", np.nanmedian(lin_reg_scores["test_r2"]))
scores_df = pd.DataFrame(lin_reg_scores)[test_metrics]
scores_df = scores_df.agg(["max", np.nanmedian]).transpose()
scores_df = aggregate_and_transpose(scores_df, statistics=["max", np.nanmedian])
scores_df["method"] = "linear_reg"
scores = pd.concat([scores, scores_df])
del lin_reg
del lin_reg_scores
ridge_reg = linear_model.Ridge(alpha=0.5)
ridge_reg_scores = cross_validate(
@ -194,16 +232,18 @@ def run_all_regression_models(input_csv):
X=data_x,
y=data_y,
groups=data_groups,
cv=logo,
cv=cross_validator,
n_jobs=-1,
scoring=metrics,
)
print("Ridge regression")
scores_df = pd.DataFrame(ridge_reg_scores)[test_metrics]
scores_df = scores_df.agg(["max", np.nanmedian]).transpose()
scores_df = aggregate_and_transpose(scores_df, statistics=["max", np.nanmedian])
scores_df["method"] = "ridge_reg"
scores = pd.concat([scores, scores_df])
del ridge_reg
del ridge_reg_scores
lasso_reg = linear_model.Lasso(alpha=0.1)
lasso_reg_score = cross_validate(
@ -211,16 +251,18 @@ def run_all_regression_models(input_csv):
X=data_x,
y=data_y,
groups=data_groups,
cv=logo,
cv=cross_validator,
n_jobs=-1,
scoring=metrics,
)
print("Lasso regression")
scores_df = pd.DataFrame(lasso_reg_score)[test_metrics]
scores_df = scores_df.agg(["max", np.nanmedian]).transpose()
scores_df = aggregate_and_transpose(scores_df, statistics=["max", np.nanmedian])
scores_df["method"] = "lasso_reg"
scores = pd.concat([scores, scores_df])
del lasso_reg
del lasso_reg_score
bayesian_ridge_reg = linear_model.BayesianRidge()
bayesian_ridge_reg_score = cross_validate(
@ -228,16 +270,18 @@ def run_all_regression_models(input_csv):
X=data_x,
y=data_y,
groups=data_groups,
cv=logo,
cv=cross_validator,
n_jobs=-1,
scoring=metrics,
)
print("Bayesian Ridge")
scores_df = pd.DataFrame(bayesian_ridge_reg_score)[test_metrics]
scores_df = scores_df.agg(["max", np.nanmedian]).transpose()
scores_df = aggregate_and_transpose(scores_df, statistics=["max", np.nanmedian])
scores_df["method"] = "bayesian_ridge"
scores = pd.concat([scores, scores_df])
del bayesian_ridge_reg
del bayesian_ridge_reg_score
ransac_reg = linear_model.RANSACRegressor()
ransac_reg_score = cross_validate(
@ -245,27 +289,37 @@ def run_all_regression_models(input_csv):
X=data_x,
y=data_y,
groups=data_groups,
cv=logo,
cv=cross_validator,
n_jobs=-1,
scoring=metrics,
)
print("RANSAC (outlier robust regression)")
scores_df = pd.DataFrame(ransac_reg_score)[test_metrics]
scores_df = scores_df.agg(["max", np.nanmedian]).transpose()
scores_df = aggregate_and_transpose(scores_df, statistics=["max", np.nanmedian])
scores_df["method"] = "RANSAC"
scores = pd.concat([scores, scores_df])
del ransac_reg
del ransac_reg_score
svr = svm.SVR()
svr_score = cross_validate(
svr, X=data_x, y=data_y, groups=data_groups, cv=logo, n_jobs=-1, scoring=metrics
svr,
X=data_x,
y=data_y,
groups=data_groups,
cv=cross_validator,
n_jobs=-1,
scoring=metrics,
)
print("Support vector regression")
scores_df = pd.DataFrame(svr_score)[test_metrics]
scores_df = scores_df.agg(["max", np.nanmedian]).transpose()
scores_df = aggregate_and_transpose(scores_df, statistics=["max", np.nanmedian])
scores_df["method"] = "SVR"
scores = pd.concat([scores, scores_df])
del svr
del svr_score
kridge = kernel_ridge.KernelRidge()
kridge_score = cross_validate(
@ -273,69 +327,130 @@ def run_all_regression_models(input_csv):
X=data_x,
y=data_y,
groups=data_groups,
cv=logo,
cv=cross_validator,
n_jobs=-1,
scoring=metrics,
)
print("Kernel Ridge regression")
scores_df = pd.DataFrame(kridge_score)[test_metrics]
scores_df = scores_df.agg(["max", np.nanmedian]).transpose()
scores_df = aggregate_and_transpose(scores_df, statistics=["max", np.nanmedian])
scores_df["method"] = "kernel_ridge"
scores = pd.concat([scores, scores_df])
del kridge
del kridge_score
gpr = gaussian_process.GaussianProcessRegressor()
gpr_score = cross_validate(
gpr, X=data_x, y=data_y, groups=data_groups, cv=logo, n_jobs=-1, scoring=metrics
gpr,
X=data_x,
y=data_y,
groups=data_groups,
cv=cross_validator,
n_jobs=-1,
scoring=metrics,
)
print("Gaussian Process Regression")
scores_df = pd.DataFrame(gpr_score)[test_metrics]
scores_df = scores_df.agg(["max", np.nanmedian]).transpose()
scores_df = aggregate_and_transpose(scores_df, statistics=["max", np.nanmedian])
scores_df["method"] = "gaussian_proc"
scores = pd.concat([scores, scores_df])
del gpr
del gpr_score
rfr = ensemble.RandomForestRegressor(max_features=0.3, n_jobs=-1)
rfr_score = cross_validate(
rfr, X=data_x, y=data_y, groups=data_groups, cv=logo, n_jobs=-1, scoring=metrics
rfr,
X=data_x,
y=data_y,
groups=data_groups,
cv=cross_validator,
n_jobs=-1,
scoring=metrics,
)
print("Random Forest Regression")
scores_df = pd.DataFrame(rfr_score)[test_metrics]
scores_df = scores_df.agg(["max", np.nanmedian]).transpose()
scores_df = aggregate_and_transpose(scores_df, statistics=["max", np.nanmedian])
scores_df["method"] = "random_forest"
scores = pd.concat([scores, scores_df])
del rfr
del rfr_score
xgb = XGBRegressor()
xgb_score = cross_validate(
xgb, X=data_x, y=data_y, groups=data_groups, cv=logo, n_jobs=-1, scoring=metrics
xgb,
X=data_x,
y=data_y,
groups=data_groups,
cv=cross_validator,
n_jobs=-1,
scoring=metrics,
)
print("XGBoost Regressor")
scores_df = pd.DataFrame(xgb_score)[test_metrics]
scores_df = scores_df.agg(["max", np.nanmedian]).transpose()
scores_df = aggregate_and_transpose(scores_df, statistics=["max", np.nanmedian])
scores_df["method"] = "XGBoost"
scores = pd.concat([scores, scores_df])
del xgb
del xgb_score
ada = ensemble.AdaBoostRegressor()
ada_score = cross_validate(
ada, X=data_x, y=data_y, groups=data_groups, cv=logo, n_jobs=-1, scoring=metrics
ada,
X=data_x,
y=data_y,
groups=data_groups,
cv=cross_validator,
n_jobs=-1,
scoring=metrics,
)
print("ADA Boost Regressor")
scores_df = pd.DataFrame(ada_score)[test_metrics]
scores_df = scores_df.agg(["max", np.nanmedian]).transpose()
scores_df = aggregate_and_transpose(scores_df, statistics=["max", np.nanmedian])
scores_df["method"] = "ADA_boost"
scores = pd.concat([scores, scores_df])
del ada
del ada_score
return scores
def run_all_classification_models(data_x, data_y, data_groups, cv_method):
metrics = ["accuracy", "average_precision", "recall", "f1"]
def confusion_matrix_scorer(clf, X, y):
y_pred = clf.predict(X)
cm = confusion_matrix(y, y_pred)
return {"tn": cm[0, 0], "fp": cm[0, 1], "fn": cm[1, 0], "tp": cm[1, 1]}
def aggregate_confusion_matrix(scores_dict: dict) -> pd.DataFrame:
scores_aggregated = aggregate_and_transpose(
pd.DataFrame(scores_dict), statistics=["sum"]
)
return scores_aggregated[
~scores_aggregated.test_metric.isin(["fit_time", "score_time"])
]
def run_all_classification_models(
data_x: pd.DataFrame,
data_y: pd.DataFrame,
data_groups: pd.DataFrame,
cross_validator: BaseCrossValidator,
):
data_y_value_counts = data_y.value_counts()
if len(data_y_value_counts) == 1:
raise (ValueError("There is only one unique value in data_y."))
if len(data_y_value_counts) == 2:
metrics = ["accuracy", "average_precision", "recall", "f1"]
else:
metrics = ["accuracy", "precision_micro", "recall_micro", "f1_micro"]
test_metrics = ["test_" + metric for metric in metrics]
scores = pd.DataFrame(columns=["method", "max", "mean"])
scores = pd.DataFrame(columns=["method", "test_metric", "max", "mean"])
dummy_class = DummyClassifier(strategy="most_frequent")
@ -344,17 +459,39 @@ def run_all_classification_models(data_x, data_y, data_groups, cv_method):
X=data_x,
y=data_y,
groups=data_groups,
cv=cv_method,
cv=cross_validator,
n_jobs=-1,
error_score="raise",
scoring=metrics,
)
dummy_confusion_matrix = cross_validate(
dummy_class,
X=data_x,
y=data_y,
groups=data_groups,
cv=cross_validator,
n_jobs=-1,
error_score="raise",
scoring=confusion_matrix_scorer,
)
print("Dummy")
scores_df = pd.DataFrame(dummy_score)[test_metrics]
scores_df = scores_df.agg(["max", "mean"]).transpose()
scores_df["method"] = "Dummy"
scores_df = aggregate_and_transpose(scores_df, statistics=["max", "mean"])
scores_df = pd.concat(
[
scores_df,
aggregate_confusion_matrix(dummy_confusion_matrix).rename(
columns={"sum": "mean"}
# Note: the column is misleadingly renamed to get concise output.
),
]
)
scores_df["method"] = "dummy_classifier"
scores = pd.concat([scores, scores_df])
del dummy_class
del dummy_score
del dummy_confusion_matrix
logistic_regression = linear_model.LogisticRegression()
@ -363,16 +500,37 @@ def run_all_classification_models(data_x, data_y, data_groups, cv_method):
X=data_x,
y=data_y,
groups=data_groups,
cv=cv_method,
cv=cross_validator,
n_jobs=-1,
scoring=metrics,
)
log_reg_confusion_matrix = cross_validate(
logistic_regression,
X=data_x,
y=data_y,
groups=data_groups,
cv=cross_validator,
n_jobs=-1,
scoring=confusion_matrix_scorer,
)
print("Logistic regression")
scores_df = pd.DataFrame(log_reg_scores)[test_metrics]
scores_df = scores_df.agg(["max", "mean"]).transpose()
scores_df["method"] = "logistic_reg"
scores_df = aggregate_and_transpose(scores_df, statistics=["max", "mean"])
scores_df = pd.concat(
[
scores_df,
aggregate_confusion_matrix(log_reg_confusion_matrix).rename(
columns={"sum": "mean"}
# Note: the column is misleadingly renamed to get concise output.
),
]
)
scores_df["method"] = "logistic_regression"
scores = pd.concat([scores, scores_df])
del logistic_regression
del log_reg_scores
del log_reg_confusion_matrix
svc = svm.SVC()
@ -381,16 +539,37 @@ def run_all_classification_models(data_x, data_y, data_groups, cv_method):
X=data_x,
y=data_y,
groups=data_groups,
cv=cv_method,
cv=cross_validator,
n_jobs=-1,
scoring=metrics,
)
svc_confusion_matrix = cross_validate(
svc,
X=data_x,
y=data_y,
groups=data_groups,
cv=cross_validator,
n_jobs=-1,
scoring=confusion_matrix_scorer,
)
print("Support Vector Machine")
scores_df = pd.DataFrame(svc_scores)[test_metrics]
scores_df = scores_df.agg(["max", "mean"]).transpose()
scores_df["method"] = "svc"
scores_df = aggregate_and_transpose(scores_df, statistics=["max", "mean"])
scores_df = pd.concat(
[
scores_df,
aggregate_confusion_matrix(svc_confusion_matrix).rename(
columns={"sum": "mean"}
# Note: the column is misleadingly renamed to get concise output.
),
]
)
scores_df["method"] = "SVC"
scores = pd.concat([scores, scores_df])
del svc
del svc_scores
del svc_confusion_matrix
gaussian_nb = naive_bayes.GaussianNB()
@ -399,16 +578,37 @@ def run_all_classification_models(data_x, data_y, data_groups, cv_method):
X=data_x,
y=data_y,
groups=data_groups,
cv=cv_method,
cv=cross_validator,
n_jobs=-1,
scoring=metrics,
)
gaussian_nb_confusion_matrix = cross_validate(
gaussian_nb,
X=data_x,
y=data_y,
groups=data_groups,
cv=cross_validator,
n_jobs=-1,
scoring=confusion_matrix_scorer,
)
print("Gaussian Naive Bayes")
scores_df = pd.DataFrame(gaussian_nb_scores)[test_metrics]
scores_df = scores_df.agg(["max", "mean"]).transpose()
scores_df = aggregate_and_transpose(scores_df, statistics=["max", "mean"])
scores_df = pd.concat(
[
scores_df,
aggregate_confusion_matrix(gaussian_nb_confusion_matrix).rename(
columns={"sum": "mean"}
# Note: the column is misleadingly renamed to get concise output.
),
]
)
scores_df["method"] = "gaussian_naive_bayes"
scores = pd.concat([scores, scores_df])
del gaussian_nb
del gaussian_nb_scores
del gaussian_nb_confusion_matrix
sgdc = linear_model.SGDClassifier()
@ -417,16 +617,37 @@ def run_all_classification_models(data_x, data_y, data_groups, cv_method):
X=data_x,
y=data_y,
groups=data_groups,
cv=cv_method,
cv=cross_validator,
n_jobs=-1,
scoring=metrics,
)
sgdc_confusion_matrix = cross_validate(
sgdc,
X=data_x,
y=data_y,
groups=data_groups,
cv=cross_validator,
n_jobs=-1,
scoring=confusion_matrix_scorer,
)
print("Stochastic Gradient Descent")
scores_df = pd.DataFrame(sgdc_scores)[test_metrics]
scores_df = scores_df.agg(["max", "mean"]).transpose()
scores_df["method"] = "stochastic_gradient_descent"
scores_df = aggregate_and_transpose(scores_df, statistics=["max", "mean"])
scores_df = pd.concat(
[
scores_df,
aggregate_confusion_matrix(sgdc_confusion_matrix).rename(
columns={"sum": "mean"}
# Note: the column is misleadingly renamed to get concise output.
),
]
)
scores_df["method"] = "stochastic_gradient_descent_classifier"
scores = pd.concat([scores, scores_df])
del sgdc
del sgdc_scores
del sgdc_confusion_matrix
rfc = ensemble.RandomForestClassifier()
@ -435,16 +656,37 @@ def run_all_classification_models(data_x, data_y, data_groups, cv_method):
X=data_x,
y=data_y,
groups=data_groups,
cv=cv_method,
cv=cross_validator,
n_jobs=-1,
scoring=metrics,
)
rfc_confusion_matrix = cross_validate(
rfc,
X=data_x,
y=data_y,
groups=data_groups,
cv=cross_validator,
n_jobs=-1,
scoring=confusion_matrix_scorer,
)
print("Random Forest")
scores_df = pd.DataFrame(rfc_scores)[test_metrics]
scores_df = scores_df.agg(["max", "mean"]).transpose()
scores_df["method"] = "random_forest"
scores_df = aggregate_and_transpose(scores_df, statistics=["max", "mean"])
scores_df = pd.concat(
[
scores_df,
aggregate_confusion_matrix(rfc_confusion_matrix).rename(
columns={"sum": "mean"}
# Note: the column is misleadingly renamed to get concise output.
),
]
)
scores_df["method"] = "random_forest_classifier"
scores = pd.concat([scores, scores_df])
del rfc
del rfc_scores
del rfc_confusion_matrix
xgb_classifier = XGBClassifier()
@ -453,15 +695,36 @@ def run_all_classification_models(data_x, data_y, data_groups, cv_method):
X=data_x,
y=data_y,
groups=data_groups,
cv=cv_method,
cv=cross_validator,
n_jobs=-1,
scoring=metrics,
)
xgb_confusion_matrix = cross_validate(
xgb_classifier,
X=data_x,
y=data_y,
groups=data_groups,
cv=cross_validator,
n_jobs=-1,
scoring=confusion_matrix_scorer,
)
print("XGBoost")
scores_df = pd.DataFrame(xgb_scores)[test_metrics]
scores_df = scores_df.agg(["max", "mean"]).transpose()
scores_df["method"] = "xgboost"
scores_df = aggregate_and_transpose(scores_df, statistics=["max", "mean"])
scores_df = pd.concat(
[
scores_df,
aggregate_confusion_matrix(xgb_confusion_matrix).rename(
columns={"sum": "mean"}
# Note: the column is misleadingly renamed to get concise output.
),
]
)
scores_df["method"] = "XGBoost_classifier"
scores = pd.concat([scores, scores_df])
del xgb_classifier
del xgb_scores
del xgb_confusion_matrix
return scores

View File

@ -33,7 +33,7 @@ class Preprocessing:
Args:
categorical_features (DataFrame): DataFrame including only categorical columns.
numerical_features (_type_): DataFrame including only numerical columns.
mode (int): Mode of the column with which DataFrame is filled. TODO: check mode results
mode (int): Mode of the column with which DataFrame is filled.
Returns:
DataFrame: Hot-One Encoded DataFrame.
@ -46,7 +46,7 @@ class Preprocessing:
if not categorical_features.empty:
categorical_features = pd.get_dummies(categorical_features)
return pd.concat([numerical_features, categorical_features], axis=1)
return pd.concat([numerical_features, categorical_features], axis=1), categorical_features.columns.tolist()
def one_hot_encode_train_and_test_sets(self, categorical_columns=["gender", "startlanguage", "mostcommonactivity", "homelabel"]):
@ -68,20 +68,27 @@ class Preprocessing:
categorical_columns = [col for col in self.train_X.columns if col in categorical_columns]
# For train set
train_X_categorical_features = self.train_X[categorical_columns].copy()
train_X_numerical_features = self.train_X.drop(categorical_columns, axis=1)
mode_train_X_categorical_features = train_X_categorical_features.mode().iloc[0]
self.train_X = self.one_hot_encoder(train_X_categorical_features, train_X_numerical_features, mode_train_X_categorical_features)
self.train_X, train_cat_col_names = self.one_hot_encoder(train_X_categorical_features, train_X_numerical_features, mode_train_X_categorical_features)
encoded_categorical_features = [col for col in self.train_X.columns if col.startswith(tuple(categorical_columns))]
# For test set
test_X_categorical_features = self.test_X[categorical_columns].copy()
test_X_numerical_features = self.test_X.drop(categorical_columns, axis=1)
self.test_X = self.one_hot_encoder(test_X_categorical_features, test_X_numerical_features, mode_train_X_categorical_features)
self.test_X, test_cat_col_names = self.one_hot_encoder(test_X_categorical_features, test_X_numerical_features, mode_train_X_categorical_features)
# Create categorical columns that were not found in test set and fill them with 0
missing_cols = [col for col in train_cat_col_names if col not in test_cat_col_names]
self.test_X[missing_cols] = 0
# Sort column names alphabetically
self.train_X = self.train_X.reindex(sorted(self.train_X.columns), axis=1)
self.test_X = self.test_X.reindex(sorted(self.test_X.columns), axis=1)
def imputer(self, interval_feature_list, other_feature_list, groupby_feature="pid"):

View File

@ -1,29 +0,0 @@
method,metric,max,mean
Dummy,test_accuracy,0.8557046979865772,0.8548446932649828
Dummy,test_average_precision,0.1457286432160804,0.14515530673501736
Dummy,test_recall,0.0,0.0
Dummy,test_f1,0.0,0.0
logistic_reg,test_accuracy,0.8640939597315436,0.8504895843872606
logistic_reg,test_average_precision,0.44363425265068757,0.37511495347389834
logistic_reg,test_recall,0.3023255813953488,0.24266238973536486
logistic_reg,test_f1,0.3909774436090226,0.318943511424051
svc,test_accuracy,0.8557046979865772,0.8548446932649828
svc,test_average_precision,0.44514416839823046,0.4068200938341621
svc,test_recall,0.0,0.0
svc,test_f1,0.0,0.0
gaussian_naive_bayes,test_accuracy,0.7684563758389261,0.7479123806954234
gaussian_naive_bayes,test_average_precision,0.2534828030085334,0.23379392278901853
gaussian_naive_bayes,test_recall,0.42528735632183906,0.3924619085805935
gaussian_naive_bayes,test_f1,0.34285714285714286,0.3107236284017699
stochastic_gradient_descent,test_accuracy,0.8576214405360134,0.7773610783222601
stochastic_gradient_descent,test_average_precision,0.3813093757959869,0.3617503752215592
stochastic_gradient_descent,test_recall,0.686046511627907,0.2822507350975675
stochastic_gradient_descent,test_f1,0.3652173913043478,0.21849107443075583
random_forest,test_accuracy,0.9110738255033557,0.9011129472867694
random_forest,test_average_precision,0.6998372262021191,0.6619275281099584
random_forest,test_recall,0.4069767441860465,0.35356856455493185
random_forest,test_f1,0.5691056910569107,0.5078402513053142
xgboost,test_accuracy,0.9128978224455612,0.9007711937764886
xgboost,test_average_precision,0.7366643049075349,0.698622165966308
xgboost,test_recall,0.5287356321839081,0.44346431435445066
xgboost,test_f1,0.638888888888889,0.5633957169928393
1 method metric max mean
2 Dummy test_accuracy 0.8557046979865772 0.8548446932649828
3 Dummy test_average_precision 0.1457286432160804 0.14515530673501736
4 Dummy test_recall 0.0 0.0
5 Dummy test_f1 0.0 0.0
6 logistic_reg test_accuracy 0.8640939597315436 0.8504895843872606
7 logistic_reg test_average_precision 0.44363425265068757 0.37511495347389834
8 logistic_reg test_recall 0.3023255813953488 0.24266238973536486
9 logistic_reg test_f1 0.3909774436090226 0.318943511424051
10 svc test_accuracy 0.8557046979865772 0.8548446932649828
11 svc test_average_precision 0.44514416839823046 0.4068200938341621
12 svc test_recall 0.0 0.0
13 svc test_f1 0.0 0.0
14 gaussian_naive_bayes test_accuracy 0.7684563758389261 0.7479123806954234
15 gaussian_naive_bayes test_average_precision 0.2534828030085334 0.23379392278901853
16 gaussian_naive_bayes test_recall 0.42528735632183906 0.3924619085805935
17 gaussian_naive_bayes test_f1 0.34285714285714286 0.3107236284017699
18 stochastic_gradient_descent test_accuracy 0.8576214405360134 0.7773610783222601
19 stochastic_gradient_descent test_average_precision 0.3813093757959869 0.3617503752215592
20 stochastic_gradient_descent test_recall 0.686046511627907 0.2822507350975675
21 stochastic_gradient_descent test_f1 0.3652173913043478 0.21849107443075583
22 random_forest test_accuracy 0.9110738255033557 0.9011129472867694
23 random_forest test_average_precision 0.6998372262021191 0.6619275281099584
24 random_forest test_recall 0.4069767441860465 0.35356856455493185
25 random_forest test_f1 0.5691056910569107 0.5078402513053142
26 xgboost test_accuracy 0.9128978224455612 0.9007711937764886
27 xgboost test_average_precision 0.7366643049075349 0.698622165966308
28 xgboost test_recall 0.5287356321839081 0.44346431435445066
29 xgboost test_f1 0.638888888888889 0.5633957169928393

View File

@ -1,29 +0,0 @@
method,metric,max,mean
Dummy,test_accuracy,1.0,0.8524114578096439
Dummy,test_average_precision,0.7,0.14758854219035614
Dummy,test_recall,0.0,0.0
Dummy,test_f1,0.0,0.0
logistic_reg,test_accuracy,0.9824561403508771,0.8445351955631311
logistic_reg,test_average_precision,1.0,0.44605167668563583
logistic_reg,test_recall,1.0,0.25353566685532386
logistic_reg,test_f1,0.823529411764706,0.27951926390778625
svc,test_accuracy,1.0,0.8524114578096439
svc,test_average_precision,0.9612401707068228,0.44179454944271934
svc,test_recall,0.0,0.0
svc,test_f1,0.0,0.0
gaussian_naive_bayes,test_accuracy,0.9,0.7491301746887129
gaussian_naive_bayes,test_average_precision,0.9189430193277607,0.2833170327386991
gaussian_naive_bayes,test_recall,1.0,0.3743761174081108
gaussian_naive_bayes,test_f1,0.7000000000000001,0.2698456659235668
stochastic_gradient_descent,test_accuracy,1.0,0.7926428596764739
stochastic_gradient_descent,test_average_precision,1.0,0.4421948838597582
stochastic_gradient_descent,test_recall,1.0,0.30156420704502945
stochastic_gradient_descent,test_f1,0.8148148148148148,0.24088393234361388
random_forest,test_accuracy,1.0,0.8722158105763481
random_forest,test_average_precision,1.0,0.49817066323226833
random_forest,test_recall,1.0,0.18161263127840668
random_forest,test_f1,1.0,0.2508096532365307
xgboost,test_accuracy,1.0,0.8812627400277729
xgboost,test_average_precision,1.0,0.5505695112208401
xgboost,test_recall,1.0,0.2896161238315027
xgboost,test_f1,0.9411764705882353,0.36887408735855665
1 method metric max mean
2 Dummy test_accuracy 1.0 0.8524114578096439
3 Dummy test_average_precision 0.7 0.14758854219035614
4 Dummy test_recall 0.0 0.0
5 Dummy test_f1 0.0 0.0
6 logistic_reg test_accuracy 0.9824561403508771 0.8445351955631311
7 logistic_reg test_average_precision 1.0 0.44605167668563583
8 logistic_reg test_recall 1.0 0.25353566685532386
9 logistic_reg test_f1 0.823529411764706 0.27951926390778625
10 svc test_accuracy 1.0 0.8524114578096439
11 svc test_average_precision 0.9612401707068228 0.44179454944271934
12 svc test_recall 0.0 0.0
13 svc test_f1 0.0 0.0
14 gaussian_naive_bayes test_accuracy 0.9 0.7491301746887129
15 gaussian_naive_bayes test_average_precision 0.9189430193277607 0.2833170327386991
16 gaussian_naive_bayes test_recall 1.0 0.3743761174081108
17 gaussian_naive_bayes test_f1 0.7000000000000001 0.2698456659235668
18 stochastic_gradient_descent test_accuracy 1.0 0.7926428596764739
19 stochastic_gradient_descent test_average_precision 1.0 0.4421948838597582
20 stochastic_gradient_descent test_recall 1.0 0.30156420704502945
21 stochastic_gradient_descent test_f1 0.8148148148148148 0.24088393234361388
22 random_forest test_accuracy 1.0 0.8722158105763481
23 random_forest test_average_precision 1.0 0.49817066323226833
24 random_forest test_recall 1.0 0.18161263127840668
25 random_forest test_f1 1.0 0.2508096532365307
26 xgboost test_accuracy 1.0 0.8812627400277729
27 xgboost test_average_precision 1.0 0.5505695112208401
28 xgboost test_recall 1.0 0.2896161238315027
29 xgboost test_f1 0.9411764705882353 0.36887408735855665

7
pyproject.toml 100644
View File

@ -0,0 +1,7 @@
[tool.isort]
profile = "black"
py_version = 311
skip_gitignore = "true"
[tool.black]
target-version = ["py311"]

2
rapids

@ -1 +1 @@
Subproject commit 63f5a526fce4d288499168e1701adadb8b885d82
Subproject commit 059774bda10545a83ab282f59eb7a329fef9ee4c

View File

@ -1,8 +1,7 @@
import os
import sqlalchemy.engine.url
from dotenv import load_dotenv
from sqlalchemy import create_engine
from sqlalchemy import URL, create_engine
from sqlalchemy.orm import sessionmaker
load_dotenv()
@ -11,7 +10,7 @@ testing: bool = False
db_password = os.getenv("DB_PASSWORD")
db_uri = sqlalchemy.engine.url.URL(
db_uri = URL.create(
drivername="postgresql+psycopg2",
username="staw_db",
password=db_password,

View File

@ -0,0 +1,60 @@
---
title: "Reliability of SAM threat and challenge and COPE"
output: html_notebook
---
```{r libraries, message=FALSE, warning=FALSE, include=FALSE, cache=FALSE}
library(conflicted)
library(here)
library(tidyverse)
library(magrittr)
library(lavaan)
library(kableExtra)
conflicts_prefer(
readr::col_factor,
purrr::discard,
dplyr::filter,
dplyr::lag,
purrr::set_names,
tidyr::extract,
kableExtra::group_rows
)
```
```{r style, include=FALSE, cache=FALSE}
styler::style_file(
here("statistical_analysis", "scale_reliability.Rmd"),
scope = "tokens",
indent_by = 4L
)
```
The data were preprocessed and cleaned using [expl_esm_labels.py](../exploration/expl_esm_labels.py) script and read as csv here.
```{r read_data}
COL_TYPES <- cols(
.default = col_double(),
participant_id = col_factor(),
username = col_factor(),
device_id = col_factor(),
esm_trigger = col_factor(),
esm_instructions = col_factor(),
double_esm_user_answer_timestamp = col_double(),
datetime_lj = col_datetime(format = ""),
date_lj = col_date(format = ""),
time = col_factor(),
esm_user_answer = col_factor()
)
df_SAM <- read_csv(here("data", "raw", "df_esm_SAM_threat_challenge.csv"), col_types = COL_TYPES)
df_COPE <- read_csv(here("data", "raw", "df_esm_COPE.csv"), col_types = COL_TYPES)
```
Demonstrate factor analysis for a single participant.
```{r}
df_COPE %>%
group_by(question_id, questionnaire_id) %>%
count()
```

View File

@ -0,0 +1,20 @@
Version: 1.0
RestoreWorkspace: Default
SaveWorkspace: Default
AlwaysSaveHistory: Default
EnableCodeIndexing: Yes
UseSpacesForTab: No
NumSpacesForTab: 4
Encoding: UTF-8
RnwWeave: Sweave
LaTeX: XeLaTeX
AutoAppendNewline: Yes
StripTrailingWhitespace: Yes
PythonType: conda
PythonVersion: 3.11.3
PythonPath: E:/ProgramData/mambaforge/envs/straw2analysis/python.exe