rapids/rules/preprocessing.snakefile

141 lines
6.2 KiB
Plaintext
Raw Normal View History

rule download_participants:
params:
group = config["DOWNLOAD_PARTICIPANTS"]["GROUP"],
2020-02-18 22:41:10 +01:00
ignored_device_ids = config["DOWNLOAD_PARTICIPANTS"]["IGNORED_DEVICE_IDS"],
timezone = config["TIMEZONE"]
priority: 1
script:
"../src/data/download_participants.R"
2019-10-24 18:11:24 +02:00
rule download_dataset:
input:
"data/external/{pid}"
params:
group = config["DOWNLOAD_DATASET"]["GROUP"],
2020-02-20 21:51:22 +01:00
table = "{sensor}",
timezone = config["TIMEZONE"],
aware_multiplatform_tables = config["ACTIVITY_RECOGNITION"]["DB_TABLE"]["ANDROID"] + "," + config["ACTIVITY_RECOGNITION"]["DB_TABLE"]["IOS"] + "," + config["CONVERSATION"]["DB_TABLE"]["ANDROID"] + "," + config["CONVERSATION"]["DB_TABLE"]["IOS"],
unifiable_sensors = {"calls": config["CALLS"]["DB_TABLE"], "battery": config["BATTERY"]["DB_TABLE"], "screen": config["SCREEN"]["DB_TABLE"], "ios_activity_recognition": config["ACTIVITY_RECOGNITION"]["DB_TABLE"]["IOS"]}
2019-10-24 18:11:24 +02:00
output:
2019-10-24 22:08:05 +02:00
"data/raw/{pid}/{sensor}_raw.csv"
2019-10-24 18:11:24 +02:00
script:
2019-10-24 22:08:05 +02:00
"../src/data/download_dataset.R"
PHONE_SENSORS = []
2020-07-24 22:35:41 +02:00
PHONE_SENSORS.extend([config["MESSAGES"]["DB_TABLE"], config["CALLS"]["DB_TABLE"], config["BARNETT_LOCATION"]["DB_TABLE"], config["DORYAB_LOCATION"]["DB_TABLE"], config["BLUETOOTH"]["DB_TABLE"], config["BATTERY"]["DB_TABLE"], config["SCREEN"]["DB_TABLE"], config["LIGHT"]["DB_TABLE"], config["ACCELEROMETER"]["DB_TABLE"], config["APPLICATIONS_FOREGROUND"]["DB_TABLE"], config["CONVERSATION"]["DB_TABLE"]["ANDROID"], config["CONVERSATION"]["DB_TABLE"]["IOS"], config["ACTIVITY_RECOGNITION"]["DB_TABLE"]["ANDROID"], config["ACTIVITY_RECOGNITION"]["DB_TABLE"]["IOS"]])
if len(config["WIFI"]["DB_TABLE"]["VISIBLE_ACCESS_POINTS"]) > 0:
PHONE_SENSORS.append(config["WIFI"]["DB_TABLE"]["VISIBLE_ACCESS_POINTS"])
2020-07-24 22:35:41 +02:00
if len(config["WIFI"]["DB_TABLE"]["CONNECTED_ACCESS_POINTS"]) > 0:
PHONE_SENSORS.append(config["WIFI"]["DB_TABLE"]["CONNECTED_ACCESS_POINTS"])
2020-07-24 22:35:41 +02:00
2019-10-24 22:08:05 +02:00
rule readable_datetime:
input:
sensor_input = rules.download_dataset.output
params:
timezones = None,
fixed_timezone = config["READABLE_DATETIME"]["FIXED_TIMEZONE"]
2020-01-15 23:18:10 +01:00
wildcard_constraints:
sensor = '.*(' + '|'.join([re.escape(x) for x in PHONE_SENSORS]) + ').*' # only process smartphone sensors, not fitbit
2019-10-24 22:08:05 +02:00
output:
"data/raw/{pid}/{sensor}_with_datetime.csv"
script:
2019-11-05 18:34:22 +01:00
"../src/data/readable_datetime.R"
2020-07-09 19:01:50 +02:00
rule phone_sensed_bins:
2019-11-05 18:34:22 +01:00
input:
2020-07-09 19:01:50 +02:00
all_sensors = expand("data/raw/{{pid}}/{sensor}_with_datetime.csv", sensor=config["PHONE_VALID_SENSED_BINS"]["TABLES"])
2019-11-05 18:34:22 +01:00
params:
2020-07-09 19:01:50 +02:00
bin_size = config["PHONE_VALID_SENSED_BINS"]["BIN_SIZE"]
2019-11-05 18:34:22 +01:00
output:
2020-07-09 19:01:50 +02:00
"data/interim/{pid}/phone_sensed_bins.csv"
2019-11-05 18:34:22 +01:00
script:
2020-07-09 19:01:50 +02:00
"../src/data/phone_sensed_bins.R"
2020-07-09 19:01:50 +02:00
rule phone_valid_sensed_days:
input:
2020-07-09 19:01:50 +02:00
phone_sensed_bins = "data/interim/{pid}/phone_sensed_bins.csv"
params:
min_valid_hours_per_day = "{min_valid_hours_per_day}",
min_valid_bins_per_hour = "{min_valid_bins_per_hour}"
output:
"data/interim/{pid}/phone_valid_sensed_days_{min_valid_hours_per_day}hours_{min_valid_bins_per_hour}bins.csv"
script:
2020-07-09 19:01:50 +02:00
"../src/data/phone_valid_sensed_days.R"
rule unify_ios_android:
input:
sensor_data = "data/raw/{pid}/{sensor}_with_datetime.csv",
participant_info = "data/external/{pid}"
params:
sensor = "{sensor}",
unifiable_sensors = {"calls": config["CALLS"]["DB_TABLE"], "battery": config["BATTERY"]["DB_TABLE"], "screen": config["SCREEN"]["DB_TABLE"], "ios_activity_recognition": config["ACTIVITY_RECOGNITION"]["DB_TABLE"]["IOS"]}
output:
"data/raw/{pid}/{sensor}_with_datetime_unified.csv"
script:
2019-12-10 00:23:00 +01:00
"../src/data/unify_ios_android.R"
rule resample_fused_location:
input:
locations = "data/raw/{pid}/{sensor}_raw.csv",
2019-12-10 00:23:00 +01:00
phone_sensed_bins = rules.phone_sensed_bins.output
params:
2020-07-09 19:01:50 +02:00
bin_size = config["PHONE_VALID_SENSED_BINS"]["BIN_SIZE"],
2019-12-10 00:23:00 +01:00
timezone = config["RESAMPLE_FUSED_LOCATION"]["TIMEZONE"],
consecutive_threshold = config["RESAMPLE_FUSED_LOCATION"]["CONSECUTIVE_THRESHOLD"],
time_since_valid_location = config["RESAMPLE_FUSED_LOCATION"]["TIME_SINCE_VALID_LOCATION"]
output:
"data/raw/{pid}/{sensor}_resampled.csv"
2019-12-10 00:23:00 +01:00
script:
2020-01-15 23:18:10 +01:00
"../src/data/resample_fused_location.R"
rule application_genres:
input:
"data/raw/{pid}/{sensor}_with_datetime.csv"
params:
catalogue_source = config["APPLICATION_GENRES"]["CATALOGUE_SOURCE"],
catalogue_file = config["APPLICATION_GENRES"]["CATALOGUE_FILE"],
update_catalogue_file = config["APPLICATION_GENRES"]["UPDATE_CATALOGUE_FILE"],
scrape_missing_genres = config["APPLICATION_GENRES"]["SCRAPE_MISSING_GENRES"]
output:
"data/interim/{pid}/{sensor}_with_datetime_with_genre.csv"
script:
"../src/data/application_genres.R"
rule fitbit_heartrate_with_datetime:
2020-01-15 23:18:10 +01:00
input:
expand("data/raw/{{pid}}/{fitbit_table}_raw.csv", fitbit_table=config["HEARTRATE"]["DB_TABLE"])
2020-01-15 23:18:10 +01:00
params:
local_timezone = config["READABLE_DATETIME"]["FIXED_TIMEZONE"],
fitbit_sensor = "heartrate"
2020-01-15 23:18:10 +01:00
output:
summary_data = "data/raw/{pid}/fitbit_heartrate_summary_with_datetime.csv",
intraday_data = "data/raw/{pid}/fitbit_heartrate_intraday_with_datetime.csv"
2020-01-15 23:18:10 +01:00
script:
"../src/data/fitbit_readable_datetime.py"
rule fitbit_step_with_datetime:
input:
expand("data/raw/{{pid}}/{fitbit_table}_raw.csv", fitbit_table=config["STEP"]["DB_TABLE"])
params:
local_timezone = config["READABLE_DATETIME"]["FIXED_TIMEZONE"],
fitbit_sensor = "steps"
output:
intraday_data = "data/raw/{pid}/fitbit_step_intraday_with_datetime.csv"
script:
"../src/data/fitbit_readable_datetime.py"
rule fitbit_sleep_with_datetime:
input:
expand("data/raw/{{pid}}/{fitbit_table}_raw.csv", fitbit_table=config["SLEEP"]["DB_TABLE"])
params:
local_timezone = config["READABLE_DATETIME"]["FIXED_TIMEZONE"],
fitbit_sensor = "sleep"
output:
summary_data = "data/raw/{pid}/fitbit_sleep_summary_with_datetime.csv",
intraday_data = "data/raw/{pid}/fitbit_sleep_intraday_with_datetime.csv"
script:
"../src/data/fitbit_readable_datetime.py"