Why QualityClean?

Cleaning datasets is one of the most repetitive steps in every data science and machine learning workflow.

QualityClean automates common preprocessing tasks so you can focus on analysis instead of writing the same cleaning code over and over.

With a single function call, QualityClean can:

• Normalize column names
• Trim unnecessary whitespace
• Replace placeholder values with nulls
• Handle missing values
• Remove duplicate rows
• Infer and convert datatypes
• Generate detailed cleaning reports

Installation

pip install qualityclean

or

uv add qualityclean

Quick Start

import qualityclean as qc

df = qc.load("employees.csv")

result = qc.clean(df)

clean_df = result.data

qc.audit(result)

qc.export(result, "cleaned.parquet")

Example

Before

After

✔ Column names normalized

✔ Whitespace removed

✔ Missing values handled

✔ Duplicate rows removed

Features

• Fast Polars-based processing
• Automatic cleaning pipeline
• Human-readable audit reports
• CSV and Parquet support
• Detailed execution statistics
• Rule-level timing information
• Pythonic API
• Lightweight with minimal dependencies

Example Audit Report

QualityClean Report

Rows processed      : 50,000
Duplicates removed  : 1,243
Missing values fixed: 892
Whitespace trimmed  : 3,144
Execution time      : 0.18 s

API

qc.load(...)
qc.clean(...)
qc.audit(...)
qc.export(...)

Complete API documentation is available in the /docs directory.

Project Structure

qualityclean/
│
├── src/
├── tests/
├── docs/
├── README.md
└── pyproject.toml

Roadmap

• Core cleaning engine
• Built-in cleaning rules
• Audit reporting
• CSV & Parquet support
• Comprehensive test suite
• Custom rule plugins
• Performance benchmarks
• Documentation website
• PyPI release

Contributing

Contributions, bug reports, and feature requests are welcome.

Please open an issue or submit a pull request.

License

Released under the MIT License.