如何将pandas DataFrame转换为指定结构的复杂嵌套JSON字典?
Pandas多层嵌套结构转换实现
测试数据集构造
按照题目给定的三列字段(people、animal、frequencies)生成可直接复现的DataFrame:
import pandas as pd raw_rows = [ ("Bob", "dog", 2), ("Bob", "cat", 1), ("Bob", "parrot", 1), ("Kate", "dog", 1), ("Kate", "parrot", 1), ("Joy", "cat", 3), ] df = pd.DataFrame(raw_rows, columns=["people", "animal", "frequencies"])
转换核心逻辑
按照目标结构的两层嵌套关系逐层组装:
- 第一层:按
animal维度拆分数据,对应顶层children数组的每个条目 - 第二层:每个动物条目下,关联对应的人物姓名和出现频率,组装为
people子数组
直接可运行的转换代码:
children = [] # 按动物分组,sort=False可保留数据原始出现顺序 for animal, group_df in df.groupby("animal", sort=False): # 组装当前动物下的人物频率列表 people_list = [ {"name": person_row["people"], "freq": person_row["frequencies"]} for _, person_row in group_df.iterrows() ] children.append({ "name": animal, "people": people_list }) # 组装顶层结构 target_result = { "title": "frequencies", "children": children }
输出结构校验
转换完成后得到的target_result可直接通过json.dumps()序列化为目标格式,结构示意:
{ "title": "frequencies", "children": [ { "name": "dog", "people": [ {"name": "Bob", "freq": 2}, {"name": "Kate", "freq": 1} ] }, { "name": "cat", "people": [ {"name": "Bob", "freq": 1}, {"name": "Joy", "freq": 3} ] }, { "name": "parrot", "people": [ {"name": "Bob", "freq": 1}, {"name": "Kate", "freq": 1} ] } ] }
若需优化大数据量下的转换效率,可替换
iterrows的逐行遍历为向量化操作:people_list = group_df.rename( columns={"people": "name", "frequencies": "freq"} )[["name", "freq"]].to_dict("records")
内容的提问来源于stack exchange,提问作者blah
相关产品推荐
相关产品推荐

