如何使用Pandas将索引值转换为列并将嵌套字典转换为指定格式的DataFrame

Let’s tackle this two-part problem—first getting your nested dictionary into a clean, usable DataFrame, then converting index values to a dedicated column. I’ve dealt with messy nested dict structures in pandas plenty of times, so here’s a straightforward solution:

1. Convert the Nested Dictionary to the Expected DataFrame #

Your original pd.DataFrame(my_dict['RuleSet']) call creates a DataFrame with columns 0, 1, 6 (the outer keys of your RuleSet dict), where each cell holds a nested dictionary. That’s almost certainly not the row-based structure you want.

Instead, use pd.json_normalize()—it’s built specifically for flattening nested JSON/dict data. Here’s how to apply it to your case:

import pandas as pd

my_dict = {'RuleSet': {'0': {'RuleSetID': '0', 'RuleSetName': 'Allgemein', 'Rules': [{'RulesID': '10', 'RuleName': 'Gemeinde Seiten', 'GroupHits': '2', 'KeyWordGroups': ['100', '101', '102']}]}, '1': {'RuleSetID': '1', 'RuleSetName': 'Portale Berlin', 'Rules': [{'RulesID': '11', 'RuleName': 'Portale Berlin', 'GroupHits': '4', 'KeyWordGroups': ['100', '101', '102', '107']}]}, '6': {'RuleSetID': '6', 'RuleSetName': 'Zwangsvollstr. Berlin', 'Rules': [{'RulesID': '23', 'RuleName': 'Zwangsvollstr. Berlin', 'GroupHits': '1', 'KeyWordGroups': ['100', '101']}]}}}

# Extract the core RuleSet entries (ignore the outer '0', '1', '6' keys)
rule_set_data = list(my_dict['RuleSet'].values())

# Flatten the nested Rules list while retaining RuleSet metadata
rules_pd = pd.json_normalize(
    rule_set_data,
    record_path='Rules',  # Expand the nested Rules list into individual rows
    meta=['RuleSetID', 'RuleSetName']  # Keep these top-level fields as columns
)

This will give you a clean, row-oriented DataFrame with columns:
RulesID, RuleName, GroupHits, KeyWordGroups, RuleSetID, RuleSetName

Each row represents a single Rule, linked to its parent RuleSet—exactly the structured format you’re probably expecting.

2. Convert Index Values to a Column #

There are two common approaches for this, depending on whether you want to keep the original index or reset it:

Option 1: Reset the index and save old values to a new column #

Use reset_index() with the names parameter to name your new index column:

# Convert current index to a column named "original_index" and reset to default 0-based index
rules_pd = rules_pd.reset_index(names='original_index')

Option 2: Keep the original index and add it as a new column #

If you want to preserve the original index while adding its values as a column, assign it directly:

# Add the current index as a column without modifying the existing index
rules_pd['original_index'] = rules_pd.index

Choose the option that fits your workflow—both work well, depending on whether you need to retain the original index order.

内容的提问来源于stack exchange，提问作者Daniel