rapids/rules/preprocessing.snakefile

33 lines
1.1 KiB
Plaintext
Raw Normal View History

2019-10-24 18:11:24 +02:00
rule download_dataset:
input:
"data/external/{pid}"
params:
group = config["DOWNLOAD_DATASET"]["GROUP"],
2019-10-24 22:08:05 +02:00
table = "{sensor}"
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"
rule readable_datetime:
input:
sensor_input = rules.download_dataset.output
params:
timezones = None,
fixed_timezone = config["READABLE_DATETIME"]["FIXED_TIMEZONE"]
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"
rule phone_valid_sensed_days:
input:
all_sensors = expand("data/raw/{{pid}}/{sensor}_with_datetime.csv", sensor=config["SENSORS"])
params:
bin_size = config["PHONE_VALID_SENSED_DAYS"]["BIN_SIZE"],
min_valid_hours = config["PHONE_VALID_SENSED_DAYS"]["MIN_VALID_HOURS"],
min_bins_per_hour = config["PHONE_VALID_SENSED_DAYS"]["MIN_BINS_PER_HOUR"]
output:
"data/interim/{pid}/phone_valid_sensed_days.csv"
script:
"../src/data/phone_valid_sensed_days.R"