Take maximum value for lab on day.

scripts/transform_labs.py

@@ -180,6 +180,13 @@ def calculate_shift(labs, date_shift):
     return merged
 
 
+def calculate_maximum(labs):
+    '''Get maximum lab value for a given day'''
+
+    labs_max = labs.loc[labs.groupby(['mrn', 'date', 'lab'])['value'].idxmax()]
+    return labs_max
+
+
 def main():
     args = get_args()
     labs = args.file
@@ -213,8 +220,11 @@ def main():
     # Calculate Date Shift
     shifted_df = calculate_shift(labs_flag, date_shift)
 
+    # Filter and get maximum lab value for shifted date
+    max_lab = calculate_maximum(shifted_df)
+
     # Write out lab table
-    shifted_df.to_csv(lab_table, index=False)
+    max_lab.to_csv(lab_table, index=False)
 
     # Write out unmapped sites if there are any
     if unmapped_labs.shape[0] > 1: