construct_prescriptions.py

import pandas as pd
import sys
import os
import json

infile = sys.argv[1]
outfile = sys.argv[2]

df = pd.read_csv(infile, sep = "|", dtype='object')
df['Month'] = df["Incurred Date"].astype(str).str[:6].astype(int)

prescriptions = {}
p = df.loc[:,["Month",
              "Drug Code",
              "GPI",
              "Drug Name",
              "Strength",
              "Member Zip Code DOS",
              "Member County DOS",
              "Service Provider ZIP",
              "Service Provider County"]]
p.rename(columns = {"Drug Code": "NDC"}, inplace = True)
p.replace({"Unknown": None, "Unkno": None, "Unspecified": None}, inplace = True)
grouped = p.groupby(by = ["NDC"], dropna=False)

for drug, frame in grouped:
    if drug in prescriptions.keys():
        prescriptions[drug] = pd.concat([prescriptions[drug], frame], ignore_index = True)
    else:
        prescriptions[drug] = frame.reset_index(drop = True)
for key in prescriptions:
    prescriptions[key] = prescriptions[key].to_dict()
with open(outfile, "w") as f:
    json.dump(prescriptions, f)