Updated G_A_R features with epochs

Snakefile

@@ -21,7 +21,8 @@ rule all:
         expand("data/processed/{pid}/bluetooth_{segment}.csv",
                             pid=config["PIDS"], 
                             segment = config["BLUETOOTH"]["DAY_SEGMENTS"]),
-        expand("data/processed/{pid}/google_activity_recognition.csv",pid=config["PIDS"]),
+        expand("data/processed/{pid}/google_activity_recognition_{segment}.csv",pid=config["PIDS"], 
+                            segment = config["GOOGLE_ACTIVITY_RECOGNITION"]["DAY_SEGMENTS"]),
         expand("data/processed/{pid}/battery_daily.csv", pid=config["PIDS"]),
         # Reports
         expand("reports/figures/{pid}/{sensor}_heatmap_rows.html", pid=config["PIDS"], sensor=config["SENSORS"]),

config.yaml

@@ -53,3 +53,7 @@ BARNETT_LOCATION:
 BLUETOOTH:
   DAY_SEGMENTS: *day_segments
   METRICS: ["countscans", "uniquedevices", "countscansmostuniquedevice"]
+
+GOOGLE_ACTIVITY_RECOGNITION:
+  DAY_SEGMENTS: *day_segments
+  METRICS: ['count','most_common_activity','number_unique_activities','activity_change_count']

rules/features.snakefile

@@ -55,8 +55,11 @@ rule bluetooth_metrics:
 rule activity_metrics:
     input:
         "data/raw/{pid}/plugin_google_activity_recognition_with_datetime.csv"
+    params:
+        segment = "{day_segment}",
+        metrics = config["GOOGLE_ACTIVITY_RECOGNITION"]["METRICS"]
     output:
-        "data/processed/{pid}/google_activity_recognition.csv"
+        "data/processed/{pid}/google_activity_recognition_{day_segment}.csv"
     script:
         "../src/features/google_activity_recognition.py"
 

src/features/google_activity_recognition.py

@@ -2,51 +2,37 @@ import pandas as pd
 import numpy as np
 import scipy.stats as stats
 
+day_segment = snakemake.params["segment"]
+
 #Read csv into a pandas dataframe
 data = pd.read_csv(snakemake.input[0])
-column = ['local_date_time','count','most_common_activity','number_unique_activities','activity_change_count']
+columns = ['count','most_common_activity','count_unique_activities','activity_change_count']
-finalDataset = pd.DataFrame(columns=column)
+columns = list("ar_" + str(day_segment) + "_" + column for column in columns)
-finalDataset.set_index('local_date_time',inplace=True)
 
 if data.empty:
-    finalDataset.to_csv(snakemake.output[0])
+    finalDataset = pd.DataFrame(columns = columns)
-
 else:
-    #Resampling each of the required features as a pandas series
     data.local_date_time = pd.to_datetime(data.local_date_time)
     resampledData = data.set_index(data.local_date_time)
-    resampledData = resampledData[~resampledData.index.duplicated()]
-    resampledData.rename_axis('time',axis='columns',inplace=True)
     resampledData.drop(columns=['local_date_time'],inplace=True)
 
-    #Finding count grouped by day
+    if(day_segment!='daily'):
-    count = pd.DataFrame()
+        resampledData = resampledData.loc[resampledData['local_day_segment'] == str(day_segment)]
+    
     count = resampledData['activity_type'].resample('D').count()
-    count = count.rename(columns={"activity_type":"count"})
 
     #Finding most common activity of the day
-    mostCommonActivity = pd.DataFrame()
     mostCommonActivity = resampledData['activity_type'].resample('D').apply(lambda x:stats.mode(x)[0])
-    mostCommonActivity = mostCommonActivity.rename(columns={'activity_type':'most_common_activity'})
 
     #finding different number of activities during a day
-    uniqueActivities = pd.DataFrame()
-    # countChanges = resampledData.to_period('D').groupby(resampledData.index)['activity_type'].value_counts()
     uniqueActivities = resampledData['activity_type'].resample('D').nunique()
     
     #finding Number of times activity changed
-    resampledData['activity_type_shift'] = resampledData['activity_type'].shift()
+    resampledData['activity_type_shift'] = resampledData['activity_type'].shift().fillna(resampledData['activity_type'].head(1),inplace=True)
-    resampledData['activity_type_shift'].fillna(resampledData['activity_type'].head(1),inplace=True)
-    #resampledData['different_activity'] = resampledData['activity_type'].apply(lambda x: 0 if resampledData['activity_type'] == resampledData['activity_type_shift'] else 1, axis=1)
     resampledData['different_activity'] = np.where(resampledData['activity_type']!=resampledData['activity_type_shift'],1,0)
-    countChanges = pd.DataFrame()
     countChanges = resampledData['different_activity'].resample('D').sum()
-
-    #Concatenating all the processed data only, no other sensor data is added here for simplicity
-    finalDataset = pd.DataFrame()
     finalDataset = pd.concat([count, mostCommonActivity, uniqueActivities, countChanges],axis=1)
-    finalDataset.rename(columns={0:"count",1:'most_common_activity','activity_type':'number_unique_activities','different_activity':'activity_change_count'},inplace = True)
 
-    #Export final dataframe with extracted features to respective PID
+finalDataset.index.names = ['local_date']
+finalDataset.columns=columns
 finalDataset.to_csv(snakemake.output[0])
-