databeacon 0.1.0

DataBeacon — A Python library for EDA, data cleaning, feature engineering, and visualization.

databeacon

A Python library for exploratory data analysis, data cleaning, feature engineering, and visualization.

---

Installation

pip install databeacon

Or install from source:

git clone https://github.com/Adityasharma-6782/databeacon.git
cd databeacon
pip install -e ".[dev]"

---

Quick Start

import pandas as pd
from databeacon import (
    summarize, describe_numerics, describe_categoricals, correlation_matrix,
    drop_duplicates, handle_missing, remove_outliers, fix_dtypes,
    encode_categoricals, scale_features, apply_transforms,
    plot_distributions, plot_correlations, plot_missing,
)

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

# ── EDA ──────────────────────────────────────────────────────────
info = summarize(df)
print(info["shape"], info["missing"])

num_stats = describe_numerics(df)
cat_stats = describe_categoricals(df)
corr = correlation_matrix(df, method="spearman")

# ── Cleaning ─────────────────────────────────────────────────────
df = drop_duplicates(df)
df = handle_missing(df, strategy="mean")          # or "median", "mode", "drop", "constant"
df = remove_outliers(df, method="iqr")            # or "zscore"
df = fix_dtypes(df, datetime_cols=["date_col"])

# ── Feature Engineering ──────────────────────────────────────────
df = encode_categoricals(df, method="onehot")     # or "label", "ordinal"
df = scale_features(df, method="standard")        # or "minmax", "robust"
df = apply_transforms(df, columns=["price"], method="log")

# ── Visualization ────────────────────────────────────────────────
plot_distributions(df)
plot_correlations(df)
plot_missing(df)

---

Modules

databeacon.eda

Function	Description
summarize(df)	Shape, dtypes, missing counts, duplicates, memory
describe_numerics(df)	Extended stats: mean, std, skewness, kurtosis, percentiles
describe_categoricals(df)	Count, unique, top value, frequency for object/category cols
correlation_matrix(df, method, threshold)	Pearson/Spearman/Kendall correlation with optional masking

databeacon.cleaning

Function	Description
drop_duplicates(df, subset, keep)	Remove duplicate rows
handle_missing(df, strategy, fill_value, drop_threshold)	Impute or drop missing values
remove_outliers(df, method, columns)	IQR or Z-score based outlier removal
fix_dtypes(df, datetime_cols, category_threshold)	Auto-infer and fix column types

databeacon.features

Function	Description
encode_categoricals(df, method, ordinal_mapping)	One-hot, label, or ordinal encoding
scale_features(df, method)	Standard, min-max, or robust scaling
apply_transforms(df, columns, method)	Log, sqrt, square, box-cox transforms
create_interaction_features(df, column_pairs, operations)	Multiply, add, subtract, divide pairs

databeacon.viz

Function	Description
plot_distributions(df)	Histograms + KDE for numeric columns
plot_correlations(df)	Annotated correlation heatmap
plot_missing(df)	Horizontal bar chart of missing %
plot_categorical_counts(df)	Value count bar charts
generate_report(df)	Auto full-page EDA report

---

Development

# Install dev dependencies
pip install -e ".[dev]"

# Run tests
pytest tests/ -v

# Run tests with coverage
pytest tests/ --cov=databeacon --cov-report=term-missing

# Lint
ruff check databeacon/

# Format
black databeacon/ tests/

---

Contributing

1. Fork the repository
2. Create a feature branch: git checkout -b feature/my-feature
3. Commit your changes: git commit -m "Add my feature"
4. Push to the branch: git push origin feature/my-feature
5. Open a Pull Request

Please add tests for any new functionality and ensure all tests pass.

---

License

MIT — see LICENSE for details.