dataframe-inspector 0.1.2

Inspect nested JSON/dict structures in pandas DataFrame columns

dataframe-inspector

Inspect nested JSON/dict structures in pandas DataFrame columns.

Installation

pip install dataframe-inspector

Usage

import pandas as pd
from dataframe_inspector import Inspector

# DataFrame with deeply nested structure (5 levels)
df = pd.DataFrame({
    'id': [1, 2],
    'response': [
        {
            'data': {
                'organization': {
                    'department': {
                        'team': {
                            'lead': {'name': 'Alice', 'id': 101},
                            'members': [
                                {'name': 'Bob', 'role': 'engineer'},
                                {'name': 'Carol', 'role': 'designer'}
                            ]
                        },
                        'name': 'Engineering',
                        'budget': 500000
                    },
                    'name': 'Tech Division'
                },
                'timestamp': '2024-01-01'
            }
        },
        {
            'data': {
                'organization': {
                    'department': {
                        'team': {
                            'lead': {'name': 'David', 'id': 102},
                            'members': [
                                {'name': 'Eve', 'role': 'analyst'}
                            ]
                        },
                        'name': 'Sales',
                        'budget': 300000
                    },
                    'name': 'Business Division'
                },
                'timestamp': '2024-01-02'
            }
        }
    ]
})

inspector = Inspector(df)

# Get overview - identifies nested vs simple columns
inspector.overview()

Output:

================================================================================
DATAFRAME OVERVIEW
================================================================================

📊 Dimensions:
  Rows: 2
  Columns: 2

🔍 Nested Columns (1):
  Use inspect_column() to explore these:
  - response (0.0% null)

📝 Simple Columns (1):
  - id (int64, 2 unique, 0.0% null)

================================================================================

# Deep dive into nested column with increased depth
inspector.inspect_column('response', max_depth=4, sample_size=1)

Output:

============================================================
Nested Column: 'response'
============================================================

Nested structure keys found (depth ≤ 4):
  - data
  - data.organization
  - data.organization.department
  - data.organization.department.budget
  - data.organization.department.name
  - data.organization.department.team
  - data.organization.name
  - data.timestamp

Sample values (first 1):

[Row 0]:
    data:
      organization:
        department:
          team:
            lead:
              name: Alice
              id: 101
            members:
              [0]:
                name: Bob
                role: engineer
              [1]:
                name: Carol
                role: designer
          name: Engineering
          budget: 500000
        name: Tech Division
      timestamp: 2024-01-01
============================================================

See more examples in the examples/ folder.

Contributing

Issues and pull requests are welcome on GitHub.