datapreprocessorbv 0.3.0

A simple and powerful data preprocessing library for cleaning datasets

datacleanbv — Advanced Data Cleaning Library

A config-driven data preprocessing library covering validation, null handling, outlier detection, scaling, and encoding.

Install

pip install -r requirements.txt
pip install -e .

Quick Start

import pandas as pd
from datacleaner import validate, replace_nulls, zscore, standard_scaler, encode, dtypeconversion

df = pd.read_csv("data.csv")

config = {
    "fixes":   {"Price": {"method": "clip", "min": 0, "max": 10000}},
    "missing": {"Price": "median", "Category": "mode"},
    "zscore":  {"Price": {"threshold": 2.5, "action": "cap"}},
    "scaling": {"Price": "standard"},
    "encoding":{"Category": "onehot"},
}

df = validate(df, config)
df = replace_nulls(df, config)
df = zscore(df, config)
df = standard_scaler(df, config)
df = encode(df, config)
df = dtypeconversion(df)

Print usage hints

from datacleaner import structure, functionName
structure()      # prints a full config template
functionName()   # prints the recommended import block

Run tests

pytest tests/ -v

Pipeline order

validate → replace_nulls → zscore / isolation_forest → standard_scaler → encode → dtypeconversion

License

MIT