quickclean 0.1.0

One-line intelligent data cleaning for pandas DataFrames

QuickClean

One-line intelligent data cleaning for pandas DataFrames.

import quickclean as qc
df_clean = qc.clean(df)

Lighter than pyjanitor. Smarter than pandas manual.

Why QuickClean?

	QuickClean	pandas manual
Lines of code	2	30
Time (100K rows)	0.26s	0.14s
Missing % left	0.0%	0.0%
Outliers handled	yes	manual
Auto dtype fix	yes	manual

Installation

pip install quickclean

# With ML-powered imputation
pip install quickclean[smart]

# Everything
pip install quickclean[full]

Usage

import quickclean as qc

# One line
df_clean = qc.clean(df)

# With options
df_clean = qc.clean(
    df,
    strategy="smart",       # 'fast' | 'smart' | 'aggressive'
    aggressiveness=0.5,     # 0.0–1.0
    verbose=True,           # print cleaning report
    preview=False,          # return analysis dict only
)

What gets cleaned automatically

• Missing values — smart imputation (median/mode/KNN/iterative)
• Outliers — adaptive detection + cap/remove/impute
• Duplicates — global + subset aware
• Data types — inference & auto-correction
• Formatting — snake_case columns, string normalization, date parsing, numeric string conversion
• Categories — fuzzy harmonization ("Jakarta" = "JAKARTA")

Performance

Rows	Time	Throughput
10K	0.04s	273K rows/s
100K	0.24s	412K rows/s
1M	2.40s	416K rows/s

Requirements

• Python >= 3.10
• pandas >= 1.5.0
• numpy >= 1.21.0

Optional: scikit-learn (smart imputation), rapidfuzz (fuzzy matching), dateparser (date parsing)