将DataFrame分组列值转换为字典的技术实现问询

Got it, let's break down how to turn that joined DataFrame into a clean, nested dictionary grouped by your job attributes. Here's a straightforward pandas-based approach that matches exactly what you're asking for.

Step 1: Set Up the Sample Data #

First, let's recreate your example DataFrame so you can test the code directly:

import pandas as pd

data = {
    'Name': ['job1', 'job1', 'job1', 'job2', 'job2', 'job2'],
    'Desc': ['desc1', 'desc1', 'desc1', 'desc2', 'desc2', 'desc2'],
    'Group': ['group1', 'group1', 'group1', 'group1', 'group1', 'group1'],
    'ConditionType': ['in', 'in', 'out', 'in', 'in', 'in'],
    'ConditionName': ['cond1', 'cond2', 'cond1', 'cond1', 'cond2', 'cond3']
}
df = pd.DataFrame(data)

Step 2: Define a Group Processing Function #

We'll write a helper function to convert each job group into the dictionary structure you want. This function extracts the fixed job details and organizes conditions by their ConditionType:

def process_job_group(group):
    # Grab the static job details (they're the same for all rows in the group)
    job_details = {
        'Desc': group['Desc'].iloc[0],
        'Group': group['Group'].iloc[0],
        'Conditions': {}
    }
    
    # Group conditions by type and collect names into lists
    condition_groups = group.groupby('ConditionType')['ConditionName'].apply(list).to_dict()
    job_details['Conditions'].update(condition_groups)
    
    return job_details

Step 3: Run the Group Conversion #

Now group the DataFrame by Name (since each name maps to unique Desc/Group in your data) and apply the function:

final_dict = df.groupby('Name').apply(process_job_group).to_dict()

Result #

The final_dict variable will be your desired nested structure:

{
    'job1': {
        'Desc': 'desc1',
        'Group': 'group1',
        'Conditions': {'in': ['cond1', 'cond2'], 'out': ['cond1']}
    },
    'job2': {
        'Desc': 'desc2',
        'Group': 'group1',
        'Conditions': {'in': ['cond1', 'cond2', 'cond3']}
    }
}

Optional: Handle Non-Unique Name/Desc/Group Combinations #

If your data ever has cases where the same Name maps to different Desc/Group values, adjust the groupby key to include all three attributes:

final_dict = df.groupby(['Name', 'Desc', 'Group']).apply(
    lambda g: g.groupby('ConditionType')['ConditionName'].apply(list).to_dict()
).to_dict()

This will use tuples like ('job1', 'desc1', 'group1') as keys in the top-level dictionary.

内容的提问来源于stack exchange，提问作者Max Mikhaylov