This repository contains three Python scripts to perform descriptive statistical summaries on CSV datasets. Each script handles nested JSON-like columns by unpacking them automatically before summarizing, supports optional grouping, and outputs results as CSV files.
- Uses built-in
csvmodule only (no third-party dependencies). - Reads CSV files with robust UTF-8 encoding and error replacement.
- Detects and unpacks nested dictionary-like columns.
- Computes count, unique values, mean, min, max, standard deviation, and most frequent values for each column.
- Supports grouping by one or more columns.
- Processes the first 10 rows by default for quick analysis.
- Outputs summaries to
python_summaries/.
- Uses Polars for fast, memory-efficient data processing.
- Detects and unpacks nested dictionary-like columns.
- Produces the same summary statistics as the pure Python version.
- Supports grouping by columns.
- Suitable for larger datasets where speed matters.
- Outputs summaries to
polars_summaries/.
- Uses Pandas and
matplotlibfor data manipulation and optional plotting. - Detects and unpacks nested dictionary-like columns.
- Computes descriptive stats for all columns and grouped subsets.
- Offers optional histogram and bar chart generation.
- Outputs summaries to
pandas_summaries/.
Run any script using Python 3.7+:
python pure_python_descriptive.py
python polars_descriptive.py
python pandas_descriptive.py