Update labs transformation.

scripts/transform_labs.py

@@ -163,6 +163,9 @@ def calculate_shift(labs, date_shift):
     date_shift['Shift'] = pd.to_timedelta(date_shift['Shift'], unit='s')
     labs['date'] = pd.to_datetime(labs['date'])
 
+    # Get labs with more than 1 value on the same datetime
+    multiple_labs = labs.groupby(['mrn', 'date', 'lab']).filter(lambda x: len(x) > 1)
+
     # Merge data
     merged = labs.merge(date_shift, left_on='mrn', right_on='MRN', how='inner')
 
@@ -177,7 +180,7 @@ def calculate_shift(labs, date_shift):
     # Int MRN
     merged.mrn = merged.mrn.astype(int)
 
-    return merged
+    return merged, multiple_labs
 
 
 def calculate_maximum(labs):
@@ -198,6 +201,8 @@ def main():
     # Make output files
     lab_table = os.path.join(out_path + 'lab_table.csv')
     unmapped_table = os.path.join(out_path + 'unmapped_lab_table.csv')
+    multiple_labs_table = os.path.join(out_path + 'multiple_lab_table.csv')
+
 
     # Read in files
     labs_df = pd.read_csv(labs)
@@ -215,7 +220,7 @@ def main():
     labs_clean, unmapped_labs = convert_units(labs_range, unitmap_df, convert_df)
 
     # Convert value flags
-    labs_flag = convert_flags(labs_clean)
+    labs_flag, multiple_labs = convert_flags(labs_clean)
 
     # Calculate Date Shift
     shifted_df = calculate_shift(labs_flag, date_shift)
@@ -224,7 +229,7 @@ def main():
     max_lab = calculate_maximum(shifted_df)
 
     # Write out lab table
-    max_lab.to_csv(lab_table, index=False)
+    calculate_maximum.to_csv(lab_table, index=False)
 
     # Write out unmapped sites if there are any
     if unmapped_labs.shape[0] > 1: