diff --git a/Snakefile b/Snakefile index ecf3dba4..ec60d9ae 100644 --- a/Snakefile +++ b/Snakefile @@ -161,6 +161,8 @@ for provider in config["FITBIT_HEARTRATE"]["PROVIDERS"].keys(): files_to_compute.extend(expand("data/raw/{pid}/fitbit_heartrate_{fitbit_data_type}_raw.csv", pid=config["PIDS"], fitbit_data_type=(["json"] if config["FITBIT_HEARTRATE"]["TABLE_FORMAT"] == "JSON" else ["summary", "intraday"]))) files_to_compute.extend(expand("data/raw/{pid}/fitbit_heartrate_{fitbit_data_type}_parsed.csv", pid=config["PIDS"], fitbit_data_type=["summary", "intraday"])) files_to_compute.extend(expand("data/raw/{pid}/fitbit_heartrate_{fitbit_data_type}_parsed_with_datetime.csv", pid=config["PIDS"], fitbit_data_type=["summary", "intraday"])) + files_to_compute.extend(expand("data/interim/{pid}/fitbit_heartrate_features/fitbit_heartrate_{language}_{provider_key}.csv", pid=config["PIDS"], language=config["FITBIT_HEARTRATE"]["PROVIDERS"][provider]["SRC_LANGUAGE"].lower(), provider_key=provider.lower())) + files_to_compute.extend(expand("data/processed/features/{pid}/fitbit_heartrate.csv", pid=config["PIDS"])) for provider in config["FITBIT_STEPS"]["PROVIDERS"].keys(): if config["FITBIT_STEPS"]["PROVIDERS"][provider]["COMPUTE"]: diff --git a/config.yaml b/config.yaml index d98ab7c6..f4e63518 100644 --- a/config.yaml +++ b/config.yaml @@ -265,8 +265,11 @@ FITBIT_HEARTRATE: PROVIDERS: RAPIDS: COMPUTE: False - SUMMARY_FEATURES: ["restinghr"] # calories features' accuracy depend on the accuracy of the participants fitbit profile (e.g. height, weight) use these with care: ["caloriesoutofrange", "caloriesfatburn", "caloriescardio", "caloriespeak"] - INTRADAY_FEATURES: ["maxhr", "minhr", "avghr", "medianhr", "modehr", "stdhr", "diffmaxmodehr", "diffminmodehr", "entropyhr", "minutesonoutofrangezone", "minutesonfatburnzone", "minutesoncardiozone", "minutesonpeakzone"] + FEATURES: + SUMMARY: ["restinghr"] # calories features' accuracy depend on the accuracy of the participants fitbit profile (e.g. height, weight) use these with care: ["caloriesoutofrange", "caloriesfatburn", "caloriescardio", "caloriespeak"] + INTRADAY: ["maxhr", "minhr", "avghr", "medianhr", "modehr", "stdhr", "diffmaxmodehr", "diffminmodehr", "entropyhr", "minutesonoutofrangezone", "minutesonfatburnzone", "minutesoncardiozone", "minutesonpeakzone"] + SRC_FOLDER: "rapids" # inside src/features/fitbit_heartrate + SRC_LANGUAGE: "python" FITBIT_STEPS: diff --git a/rules/features.smk b/rules/features.smk index b3f965f3..10e8b353 100644 --- a/rules/features.smk +++ b/rules/features.smk @@ -216,6 +216,19 @@ rule phone_wifi_visible_r_features: script: "../src/features/entry.R" +rule fitbit_heartrate_python_features: + input: + sensor_data = expand("data/raw/{{pid}}/fitbit_heartrate_{fitbit_data_type}_parsed_with_datetime.csv", fitbit_data_type=["summary", "intraday"]), + day_segments_labels = "data/interim/day_segments/{pid}_day_segments_labels.csv" + params: + provider = lambda wildcards: config["FITBIT_HEARTRATE"]["PROVIDERS"][wildcards.provider_key.upper()], + provider_key = "{provider_key}", + sensor_key = "fitbit_heartrate" + output: + "data/interim/{pid}/fitbit_heartrate_features/fitbit_heartrate_python_{provider_key}.csv" + script: + "../src/features/entry.py" + # rule fitbit_heartrate_features: # input: # heartrate_summary_data = "data/raw/{pid}/fitbit_heartrate_summary_with_datetime.csv", diff --git a/src/features/fitbit_heartrate/fitbit_heartrate_base.py b/src/features/fitbit_heartrate/fitbit_heartrate_base.py deleted file mode 100644 index 2952fa57..00000000 --- a/src/features/fitbit_heartrate/fitbit_heartrate_base.py +++ /dev/null @@ -1,76 +0,0 @@ -import pandas as pd -from scipy.stats import entropy - -def extractHRFeaturesFromSummaryData(heartrate_summary_data, summary_features): - heartrate_summary_features = pd.DataFrame() - if "restinghr" in summary_features: - heartrate_summary_features["heartrate_daily_restinghr"] = heartrate_summary_data["heartrate_daily_restinghr"] - # calories features might be inaccurate: they depend on users' fitbit profile (weight, height, etc.) - if "caloriesoutofrange" in summary_features: - heartrate_summary_features["heartrate_daily_caloriesoutofrange"] = heartrate_summary_data["heartrate_daily_caloriesoutofrange"] - if "caloriesfatburn" in summary_features: - heartrate_summary_features["heartrate_daily_caloriesfatburn"] = heartrate_summary_data["heartrate_daily_caloriesfatburn"] - if "caloriescardio" in summary_features: - heartrate_summary_features["heartrate_daily_caloriescardio"] = heartrate_summary_data["heartrate_daily_caloriescardio"] - if "caloriespeak" in summary_features: - heartrate_summary_features["heartrate_daily_caloriespeak"] = heartrate_summary_data["heartrate_daily_caloriespeak"] - heartrate_summary_features.reset_index(inplace=True) - - return heartrate_summary_features - -def extractHRFeaturesFromIntradayData(heartrate_intraday_data, features, day_segment): - heartrate_intraday_features = pd.DataFrame(columns=["local_date"] + ["heartrate_" + day_segment + "_" + x for x in features]) - if not heartrate_intraday_data.empty: - device_id = heartrate_intraday_data["device_id"][0] - num_rows_per_minute = heartrate_intraday_data.groupby(["local_date", "local_hour", "local_minute"]).count().mean()["device_id"] - if day_segment != "daily": - heartrate_intraday_data = heartrate_intraday_data[heartrate_intraday_data["local_day_segment"] == day_segment] - - if not heartrate_intraday_data.empty: - heartrate_intraday_features = pd.DataFrame() - - # get stats of heartrate - if "maxhr" in features: - heartrate_intraday_features["heartrate_" + day_segment + "_maxhr"] = heartrate_intraday_data[["local_date", "heartrate"]].groupby(["local_date"])["heartrate"].max() - if "minhr" in features: - heartrate_intraday_features["heartrate_" + day_segment + "_minhr"] = heartrate_intraday_data[["local_date", "heartrate"]].groupby(["local_date"])["heartrate"].min() - if "avghr" in features: - heartrate_intraday_features["heartrate_" + day_segment + "_avghr"] = heartrate_intraday_data[["local_date", "heartrate"]].groupby(["local_date"])["heartrate"].mean() - if "medianhr" in features: - heartrate_intraday_features["heartrate_" + day_segment + "_medianhr"] = heartrate_intraday_data[["local_date", "heartrate"]].groupby(["local_date"])["heartrate"].median() - if "modehr" in features: - heartrate_intraday_features["heartrate_" + day_segment + "_modehr"] = heartrate_intraday_data[["local_date", "heartrate"]].groupby(["local_date"])["heartrate"].agg(lambda x: pd.Series.mode(x)[0]) - if "stdhr" in features: - heartrate_intraday_features["heartrate_" + day_segment + "_stdhr"] = heartrate_intraday_data[["local_date", "heartrate"]].groupby(["local_date"])["heartrate"].std() - if "diffmaxmodehr" in features: - heartrate_intraday_features["heartrate_" + day_segment + "_diffmaxmodehr"] = heartrate_intraday_data[["local_date", "heartrate"]].groupby(["local_date"])["heartrate"].max() - heartrate_intraday_data[["local_date", "heartrate"]].groupby(["local_date"])["heartrate"].agg(lambda x: pd.Series.mode(x)[0]) - if "diffminmodehr" in features: - heartrate_intraday_features["heartrate_" + day_segment + "_diffminmodehr"] = heartrate_intraday_data[["local_date", "heartrate"]].groupby(["local_date"])["heartrate"].agg(lambda x: pd.Series.mode(x)[0]) - heartrate_intraday_data[["local_date", "heartrate"]].groupby(["local_date"])["heartrate"].min() - if "entropyhr" in features: - heartrate_intraday_features["heartrate_" + day_segment + "_entropyhr"] = heartrate_intraday_data[["local_date", "heartrate"]].groupby(["local_date"])["heartrate"].agg(entropy) - - # get number of minutes in each heart rate zone - for feature_name in list(set(["minutesonoutofrangezone", "minutesonfatburnzone", "minutesoncardiozone", "minutesonpeakzone"]) & set(features)): - heartrate_zone = heartrate_intraday_data[heartrate_intraday_data["heartrate_zone"] == feature_name[9:-4]] - heartrate_intraday_features["heartrate_" + day_segment + "_" + feature_name] = heartrate_zone.groupby(["local_date"])["device_id"].count() / num_rows_per_minute - heartrate_intraday_features.fillna(value={"heartrate_" + day_segment + "_" + feature_name: 0}, inplace=True) - heartrate_intraday_features.reset_index(inplace=True) - - return heartrate_intraday_features - -def base_fitbit_heartrate_features(heartrate_summary_data, heartrate_intraday_data, day_segment, requested_summary_features, requested_intraday_features): - # name of the features this function can compute - base_summary_features_names = ["restinghr", "caloriesoutofrange", "caloriesfatburn", "caloriescardio", "caloriespeak"] - base_intraday_features_names = ["maxhr", "minhr", "avghr", "medianhr", "modehr", "stdhr", "diffmaxmodehr", "diffminmodehr", "entropyhr", "minutesonoutofrangezone", "minutesonfatburnzone", "minutesoncardiozone", "minutesonpeakzone"] - # the subset of requested features this function can compute - summary_features_to_compute = list(set(requested_summary_features) & set(base_summary_features_names)) - intraday_features_to_compute = list(set(requested_intraday_features) & set(base_intraday_features_names)) - - heartrate_intraday_features = extractHRFeaturesFromIntradayData(heartrate_intraday_data, intraday_features_to_compute, day_segment) - if not heartrate_summary_data.empty and day_segment == "daily" and summary_features_to_compute != []: - heartrate_summary_features = extractHRFeaturesFromSummaryData(heartrate_summary_data, summary_features_to_compute) - heartrate_features = heartrate_intraday_features.merge(heartrate_summary_features, on=["local_date"], how="outer") - else: - heartrate_features = heartrate_intraday_features - - return heartrate_features diff --git a/src/features/fitbit_heartrate/rapids/main.py b/src/features/fitbit_heartrate/rapids/main.py new file mode 100644 index 00000000..82554076 --- /dev/null +++ b/src/features/fitbit_heartrate/rapids/main.py @@ -0,0 +1,94 @@ +import pandas as pd +from scipy.stats import entropy + +def extractHRFeaturesFromSummaryData(heartrate_summary_data, summary_features): + heartrate_summary_data.set_index("local_segment", inplace=True) + heartrate_summary_features = pd.DataFrame() + if "restinghr" in summary_features: + heartrate_summary_features["heartrate_rapids_restinghr"] = heartrate_summary_data["heartrate_daily_restinghr"] + # calories features might be inaccurate: they depend on users' fitbit profile (weight, height, etc.) + if "caloriesoutofrange" in summary_features: + heartrate_summary_features["heartrate_rapids_caloriesoutofrange"] = heartrate_summary_data["heartrate_daily_caloriesoutofrange"] + if "caloriesfatburn" in summary_features: + heartrate_summary_features["heartrate_rapids_caloriesfatburn"] = heartrate_summary_data["heartrate_daily_caloriesfatburn"] + if "caloriescardio" in summary_features: + heartrate_summary_features["heartrate_rapids_caloriescardio"] = heartrate_summary_data["heartrate_daily_caloriescardio"] + if "caloriespeak" in summary_features: + heartrate_summary_features["heartrate_rapids_caloriespeak"] = heartrate_summary_data["heartrate_daily_caloriespeak"] + heartrate_summary_features.reset_index(inplace=True) + + return heartrate_summary_features + +def extractHRFeaturesFromIntradayData(heartrate_intraday_data, features, day_segment, filter_data_by_segment): + heartrate_intraday_features = pd.DataFrame(columns=["local_segment"] + ["heartrate_rapids_" + x for x in features]) + if not heartrate_intraday_data.empty: + num_rows_per_minute = heartrate_intraday_data.groupby(["local_date", "local_hour", "local_minute"]).count().mean()["device_id"] + heartrate_intraday_data = filter_data_by_segment(heartrate_intraday_data, day_segment) + + if not heartrate_intraday_data.empty: + heartrate_intraday_features = pd.DataFrame() + + # get stats of heartrate + if "maxhr" in features: + heartrate_intraday_features["heartrate_rapids_maxhr"] = heartrate_intraday_data[["local_segment", "heartrate"]].groupby(["local_segment"])["heartrate"].max() + if "minhr" in features: + heartrate_intraday_features["heartrate_rapids_minhr"] = heartrate_intraday_data[["local_segment", "heartrate"]].groupby(["local_segment"])["heartrate"].min() + if "avghr" in features: + heartrate_intraday_features["heartrate_rapids_avghr"] = heartrate_intraday_data[["local_segment", "heartrate"]].groupby(["local_segment"])["heartrate"].mean() + if "medianhr" in features: + heartrate_intraday_features["heartrate_rapids_medianhr"] = heartrate_intraday_data[["local_segment", "heartrate"]].groupby(["local_segment"])["heartrate"].median() + if "modehr" in features: + heartrate_intraday_features["heartrate_rapids_modehr"] = heartrate_intraday_data[["local_segment", "heartrate"]].groupby(["local_segment"])["heartrate"].agg(lambda x: pd.Series.mode(x)[0]) + if "stdhr" in features: + heartrate_intraday_features["heartrate_rapids_stdhr"] = heartrate_intraday_data[["local_segment", "heartrate"]].groupby(["local_segment"])["heartrate"].std() + if "diffmaxmodehr" in features: + heartrate_intraday_features["heartrate_rapids_diffmaxmodehr"] = heartrate_intraday_data[["local_segment", "heartrate"]].groupby(["local_segment"])["heartrate"].max() - heartrate_intraday_data[["local_segment", "heartrate"]].groupby(["local_segment"])["heartrate"].agg(lambda x: pd.Series.mode(x)[0]) + if "diffminmodehr" in features: + heartrate_intraday_features["heartrate_rapids_diffminmodehr"] = heartrate_intraday_data[["local_segment", "heartrate"]].groupby(["local_segment"])["heartrate"].agg(lambda x: pd.Series.mode(x)[0]) - heartrate_intraday_data[["local_segment", "heartrate"]].groupby(["local_segment"])["heartrate"].min() + if "entropyhr" in features: + heartrate_intraday_features["heartrate_rapids_entropyhr"] = heartrate_intraday_data[["local_segment", "heartrate"]].groupby(["local_segment"])["heartrate"].agg(entropy) + + # get number of minutes in each heart rate zone + for feature_name in list(set(["minutesonoutofrangezone", "minutesonfatburnzone", "minutesoncardiozone", "minutesonpeakzone"]) & set(features)): + heartrate_zone = heartrate_intraday_data[heartrate_intraday_data["heartrate_zone"] == feature_name[9:-4]] + heartrate_intraday_features["heartrate_rapids_" + feature_name] = heartrate_zone.groupby(["local_segment"])["device_id"].count() / num_rows_per_minute + heartrate_intraday_features.fillna(value={"heartrate_rapids_" + feature_name: 0}, inplace=True) + heartrate_intraday_features.reset_index(inplace=True) + + return heartrate_intraday_features + + +def rapids_features(sensor_data_files, day_segment, provider, filter_data_by_segment, *args, **kwargs): + + heartrate_summary_data = pd.read_csv(sensor_data_files["sensor_data"][0]) + heartrate_intraday_data = pd.read_csv(sensor_data_files["sensor_data"][1]) + + requested_summary_features = provider["FEATURES"]["SUMMARY"] + requested_intraday_features = provider["FEATURES"]["INTRADAY"] + # name of the features this function can compute + base_summary_features_names = ["restinghr", "caloriesoutofrange", "caloriesfatburn", "caloriescardio", "caloriespeak"] + base_intraday_features_names = ["maxhr", "minhr", "avghr", "medianhr", "modehr", "stdhr", "diffmaxmodehr", "diffminmodehr", "entropyhr", "minutesonoutofrangezone", "minutesonfatburnzone", "minutesoncardiozone", "minutesonpeakzone"] + # the subset of requested features this function can compute + summary_features_to_compute = list(set(requested_summary_features) & set(base_summary_features_names)) + intraday_features_to_compute = list(set(requested_intraday_features) & set(base_intraday_features_names)) + + heartrate_intraday_features = extractHRFeaturesFromIntradayData(heartrate_intraday_data, intraday_features_to_compute, day_segment, filter_data_by_segment) + if not heartrate_summary_data.empty and day_segment == "daily" and summary_features_to_compute != []: + # filter by segment and skipping any non-daily segment + heartrate_summary_data = filter_data_by_segment(heartrate_summary_data, "daily") + + datetime_start_regex = "[0-9]{4}[\\-|\\/][0-9]{2}[\\-|\\/][0-9]{2} 00:00:00" + datetime_end_regex = "[0-9]{4}[\\-|\\/][0-9]{2}[\\-|\\/][0-9]{2} 23:59:59" + + segment_regex = "daily#{},{}".format(datetime_start_regex, datetime_end_regex) + heartrate_summary_data = heartrate_summary_data[heartrate_summary_data["local_segment"].str.match(segment_regex)] + + # extract daily features from summary data + heartrate_summary_features = extractHRFeaturesFromSummaryData(heartrate_summary_data, summary_features_to_compute) + + # merge summary features and intraday features + heartrate_features = heartrate_intraday_features.merge(heartrate_summary_features, on=["local_segment"], how="outer") + else: + heartrate_features = heartrate_intraday_features + + return heartrate_features diff --git a/src/features/fitbit_heartrate_features.py b/src/features/fitbit_heartrate_features.py deleted file mode 100644 index 61bad1ee..00000000 --- a/src/features/fitbit_heartrate_features.py +++ /dev/null @@ -1,16 +0,0 @@ -import pandas as pd -from fitbit_heartrate.fitbit_heartrate_base import base_fitbit_heartrate_features - -heartrate_summary_data = pd.read_csv(snakemake.input["heartrate_summary_data"], index_col=["local_date"], parse_dates=["local_date"]) -heartrate_intraday_data = pd.read_csv(snakemake.input["heartrate_intraday_data"], parse_dates=["local_date_time", "local_date"]) -day_segment = snakemake.params["day_segment"] -requested_summary_features = snakemake.params["summary_features"] -requested_intraday_features = snakemake.params["intraday_features"] -heartrate_features = pd.DataFrame(columns=["local_date"]) - -heartrate_features = heartrate_features.merge(base_fitbit_heartrate_features(heartrate_summary_data, heartrate_intraday_data, day_segment, requested_summary_features, requested_intraday_features), on="local_date", how="outer") - -requested_features = requested_summary_features + requested_intraday_features if day_segment == "daily" else requested_intraday_features -assert len(requested_features) + 1 == heartrate_features.shape[1], "The number of features in the output dataframe (=" + str(heartrate_features.shape[1]) + ") does not match the expected value (=" + str(len(requested_features)) + " + 1). Verify your fitbit heartrate feature extraction functions" - -heartrate_features.to_csv(snakemake.output[0], index=False)