如何用Python将API创建客户的响应结果存入Pandas DataFrame

我需要在账单系统中创建新客户，获取包含新客户ID和创建时间戳的API响应，并将所有响应数据存入Pandas DataFrame以便后续处理。以下是我目前的代码，但不清楚如何将多字段的响应追加到DataFrame中：

new_customers = pd.DataFrame()
for customer,r in sh2_ws_df.iterrows():
    cb.Customer.create({
        "first_name": r['Billing First Name'],
        "last_name": r['Billing Last Name'],
        "email": r['Billing Email'],
        "phone": r['Billing Phone'],
        "company": r['Account Name'],
        "auto_collection": "on",
        "net_term_days": 0,
        "allow_direct_debit": 'true',
        "taxability": "taxable",
        "locale": "en",
        "cf_referral_partner": r['cb_referral_partner'],
        "cf_business_type": r['cf_business_type'],
        "billing_address" : {
            "first_name" : r['Billing First Name'],
            "last_name" : r['Billing Last Name'],
            "email": r['Billing Email'],
            "company": r['Account Name'],
            "phone": r['Billing Phone'],
            "line1" : r['Billing Address 1'],
            "line2": r['Billing Address 2'],
            "city" : r['Billing City'],
            "state" : r['Billing State'],
            "zip" : r['Billing Zip'],
            "country" : r['Billing Country']
        }
}
    new_customers.append()
    )

你的代码存在几个核心问题：未捕获API响应数据、append()方法使用错误、语法括号不匹配。下面是修正后的高效实现方式：

推荐实现：先收集响应列表再转DataFrame #

import pandas as pd

# 初始化空列表存储每个客户的API响应数据
customer_responses = []

# 遍历源数据创建客户
for _, r in sh2_ws_df.iterrows():
    # 调用API创建客户并捕获返回的响应对象
    response = cb.Customer.create({
        "first_name": r['Billing First Name'],
        "last_name": r['Billing Last Name'],
        "email": r['Billing Email'],
        "phone": r['Billing Phone'],
        "company": r['Account Name'],
        "auto_collection": "on",
        "net_term_days": 0,
        "allow_direct_debit": 'true',
        "taxability": "taxable",
        "locale": "en",
        "cf_referral_partner": r['cb_referral_partner'],
        "cf_business_type": r['cf_business_type'],
        "billing_address" : {
            "first_name" : r['Billing First Name'],
            "last_name" : r['Billing Last Name'],
            "email": r['Billing Email'],
            "company": r['Account Name'],
            "phone": r['Billing Phone'],
            "line1" : r['Billing Address 1'],
            "line2": r['Billing Address 2'],
            "city" : r['Billing City'],
            "state" : r['Billing State'],
            "zip" : r['Billing Zip'],
            "country" : r['Billing Country']
        }
    })
    
    # 将响应对象转为字典（多数API返回的对象支持to_dict()方法，若直接返回字典可跳过此步）
    customer_data = response.to_dict() if hasattr(response, 'to_dict') else dict(response)
    
    # 将单条客户数据追加到列表
    customer_responses.append(customer_data)

# 一次性将列表转为DataFrame，效率远高于循环追加
new_customers = pd.DataFrame(customer_responses)

# 若需要展开嵌套的billing_address字段为平级列，使用json_normalize
# new_customers = pd.json_normalize(customer_responses)

关键说明 #

1. 捕获API响应：必须将cb.Customer.create()的返回值赋值给变量，才能获取到新客户ID、创建时间戳等核心数据。
2. 避免循环操作DataFrame：循环中使用append()（已被Pandas弃用）会频繁生成新DataFrame，性能极低，先收集列表再批量转换是最优方案。
3. 处理嵌套字段：如果API响应包含billing_address这类嵌套结构，pd.json_normalize()可以将嵌套字段展开为平级列，方便后续分析处理。
4. 语法修正：原代码中cb.Customer.create()的大括号未闭合，修正后确保代码语法合法。

内容的提问来源于stack exchange，提问作者Mike P