Commit Graph

1378 Commits (90ee99e4b99eeb32650df06d60e43c0ee1e66a87)

Author SHA1 Message Date
primoz e1d7607de4 Extraction of additional SO features. Min/max has been changed to nsmallest/nlargest means. 2022-06-10 12:34:48 +00:00
primoz f371249b99 First order features standardization WIP 2022-06-09 13:35:15 +00:00
primoz 64e41cfa35 Second order features standardization in config.yaml. 2022-06-07 10:39:48 +00:00
primoz 2c7ac21465 Added standardization on SO features. 2022-06-06 13:51:15 +00:00
primoz 2acf6ff9fb Exception handling in case of empty ibi. Changes of the method EDA uses in main.py. Other small corrections. 2022-06-03 12:34:36 +00:00
primoz d300f0f8f0 Fixed RAPIDS bug: error when IBI.csv is empty. 2022-06-02 11:43:49 +00:00
Primoz fbf6a77dfc Small misc changes 2022-06-02 06:41:53 +00:00
Primoz 5532043b1f Patching IBI with BVP - completed. 2022-05-25 19:39:47 +00:00
Primoz bb62497ba6 Patching IBI with BVP - selecting appropriate pipeline entry point. WIP 2022-05-24 11:07:18 +00:00
Primoz 2a8f58f5c8 Patching IBI with BVP. WIP 2022-05-20 13:18:45 +00:00
Primoz 1471c86c62 Cr-features version update in rapids venv. 2022-05-13 13:37:12 +00:00
Primoz 6864cfe775 Changes after thorough testing with available data. 2022-05-13 13:35:34 +00:00
Primoz c1564f0cae Changed wrapper method calculate_feature to its newest version (for TEMP and ACC). 2022-05-11 14:21:21 +00:00
Primoz 31e36e7400 Alternating Second order and full segment features corresponding to config settings. 2022-05-11 08:50:15 +00:00
Primoz 9cf9e1fe14 Testing and modifying the code with different E4 data. 2022-05-10 11:36:49 +00:00
Primoz f62a1302dd Cr-features corrections for ACC and TEMP sensors 2022-05-09 11:01:52 +00:00
Primoz 5638367999 Implementation of the second order features. 2022-04-25 13:07:03 +00:00
Primoz 66451160e9 Calculating HRV features with IBI.csv. 2022-04-20 10:44:51 +00:00
= 8c8fe1fec7 Modifications, mostly imports, after changes in cr-features module. 2022-04-19 13:24:46 +00:00
= 075c64d1e5 HRV: changed wrapper calcFeat method with specialized one. 2022-04-14 11:51:53 +00:00
junos c05b047c2d Correct outstanding baseline feature mistake. 2022-04-13 17:05:16 +02:00
junos 53ec52a954 Disable (SOME) feature cleaning for ESM data. 2022-04-13 16:01:31 +02:00
= 3c058e4463 Add option to calculate features within windows and store it in CSV (all sensors). 2022-04-13 13:18:23 +00:00
junos 144f0d0dcf Account for missing baseline data. 2022-04-13 14:56:28 +02:00
junos ed5314aa98 Merge remote-tracking branch 'origin/master' 2022-04-12 17:27:25 +02:00
junos 11c64cfc1a Include all participants again. 2022-04-12 17:20:19 +02:00
junos a6a37c7bd9 Drop NaN targets.
This mirrors INNER join in merge_features_and_targets_for_individual_model.py:

data = pd.concat([sensor_features, targets[["target"]]], axis=1, join="inner")
2022-04-12 17:01:49 +02:00
junos 9f5edf1c2b Revert "Add a rule for model baselines."
The example was for a classification rather than regression problem.

This reverts commit 9ab0c8f289.

# Conflicts:
#	rules/models.smk
2022-04-12 16:59:42 +02:00
junos 4ad261fae5 Rename baseline features AGAIN.
Correct other mistakes.
2022-04-12 16:55:01 +02:00
= 74cf4ada1c Cr-feat window length for all empaticas sensors. 2022-04-12 14:00:44 +00:00
junos 9ab0c8f289 Add a rule for model baselines.
Add baselines and helper functions to main models dir.
2022-04-12 14:23:58 +02:00
junos 570d2eb656 Add the file for population model to Snakefile. 2022-04-12 14:11:40 +02:00
junos f5688f6154 Add a rule to merge sensor and baseline features.
And select target as before.
2022-04-08 15:42:04 +02:00
junos b1f356c3f7 Extract a function to be used elsewhere. 2022-04-08 15:36:32 +02:00
junos 7ff3dcf5fc Move and rename target variable. 2022-04-06 18:21:09 +02:00
junos 50c0defca7 Select target columns (no parsing necessary). 2022-04-06 18:16:49 +02:00
junos ac86221662 [WIP] Add a rule to parse targets.
Does nothing for now.
2022-04-06 17:47:03 +02:00
junos baa94c4c4e Correct additional error in feature file naming.
Add the final feature file to the list in Snakefile.
2022-04-06 17:29:17 +02:00
junos d2fbef5234 Merge branch 'labels' of https://repo.ijs.si/junoslukan/rapids into labels
# Conflicts:
#	src/features/phone_esm/straw/preprocess.py
2022-04-05 19:28:37 +02:00
junos d326a1b09d Include the constant directly in main.py. 2022-04-05 19:08:43 +02:00
junos 2e545e81f0 Include feature calculations for different scales. 2022-04-05 19:05:34 +02:00
junos cbc8ae4e03 Add necessary checks for empty data frames. 2022-04-05 18:58:09 +02:00
junos f50a13167e Add feature files back to Snakefile. 2022-04-05 18:37:58 +02:00
junos e84c35a36a Remove unnecessary parameters from preprocess_esm.
And correct the newly named interim file.
2022-04-05 18:36:09 +02:00
junos e2ce68f591 Defer creation of feature files to esm_features rule. 2022-04-05 18:30:04 +02:00
junos 751b04f3f4 Pass scale names to Snakemake correctly. 2022-04-05 18:14:37 +02:00
junos 99245afca3 Try a different approach for preprocessing ESMs.
It is important that this follows generic RAPIDS pattern.
In the subsequent step of calculating features,
there is an expected file and folder structure of data/interim.
See rules/common.smk/find_features_files()
2022-04-05 18:02:31 +02:00
junos ed298a9479 Implement the basic feature extraction steps. 2022-04-05 15:46:02 +02:00
= 1c42347b9b Small changes. 2022-04-04 12:19:33 +00:00
Primoz c050174ca3 Various minimal changes. 2022-03-31 09:16:00 +00:00