Python保险客户办理数据集csv 将其中的json字符串转换为dataframe格式化处理

import pandas as pd import numpy as np import json payout_data = pd.read_csv("payout_data.csv") payout_data["payload"] = payout_data["payload"].apply(lambda x: x.replace("'", "\"")) payout_data["payload"] = payout_data["payload"].apply(lambda x: x.replace("True", "\"True\"")) payout_data["payload"] = payout_data["payload"].apply(lambda x: x.replace("False", "\"False\"")) df_rows = [] for i in range(len(payout_data.index)): df_row = pd.json_normalize(json.loads(payout_data["payload"][i])) df_rows.append(df_row) df_rows = pd.concat(df_rows) df_rows = df_rows.reset_index().iloc[:, 1:] payout_data_new = pd.concat([payout_data, df_rows], axis=1) payout_data_new["reimbursements"] = payout_data_new["reimbursements"].apply(lambda x: str(x).replace("'", "\"")) payout_data_new = payout_data_new.drop("payload", axis=1) payout_data_new.to_csv("payout_data_new.csv", index=False) reimbursements = str(payout_data_new["reimbursements"].values.tolist()) reimbursements = reimbursements.replace("'", "") payout_ids = np.array(payout_data["payout_id"]) list_re = [] for i, json_re in enumerate(json.loads(reimbursements)): df = pd.json_normalize(json_re) df["payout_id"] = payout_ids[i] list_re.append(df) df_re = pd.concat(list_re) df_re = df_re.reset_index().iloc[:, 1:] df_re.to_csv("reimbursement.csv", index=False)