featurelab 0.1.2

Comprehensive feature engineering package with statistical guidance

FeatureLab

FeatureLab is a comprehensive Python package for feature engineering, offering statistical guidance and a suite of tools to streamline data preprocessing for machine learning projects.

---

Features

• Automatic Feature Type Detection: Identify numeric, categorical, datetime, and text columns in your DataFrame.
• Missing Value Visualization: Visualize missing data patterns and distributions.
• Outlier Visualization: Easily spot and analyze outliers.
• Feature Importance Plotting: Visualize feature importance scores for model interpretability.
• Correlation Matrix Heatmaps: Explore feature correlations visually.
• PCA & RFE Visualization: Understand dimensionality reduction and feature selection results.
• Memory Optimization: Reduce DataFrame memory usage efficiently.
• Datetime Feature Expansion: Extract year, month, day, and more from datetime columns.
• Categorical Distribution Plots: Visualize the distribution of categorical features.
• Duplicate Row Visualization: Detect and visualize duplicate rows.
• Easy Integration: Designed to work seamlessly with pandas DataFrames.

---

Installation

Clone the repository and install with pip:

git clone https://github.com/yourusername/featurelab.git
cd featurelab
pip install .

Or install directly if distributed on PyPI:

pip install featurelab

---

Requirements

• Python >= 3.7
• numpy >= 1.20.0
• pandas >= 1.2.0
• scipy >= 1.6.0
• scikit-learn >= 0.24.0
• matplotlib >= 3.3.0
• seaborn >= 0.11.0
• missingno >= 0.4.2

---

Usage

Python API

import pandas as pd
from featurelab.utils import FeatureUtils
from featurelab.visualizer import Visualizer

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

# Detect column types
col_types = FeatureUtils.detect_column_types(df)
print(col_types)

# Optimize memory usage
df_optimized = FeatureUtils.memory_optimize(df)

# Visualize missing values
viz = Visualizer()
viz.plot_null_matrix(df)

# Plot feature importance (example)
# importance_scores = ... # pd.Series with feature importances
# viz.plot_feature_importance(importance_scores)

CLI (if implemented)

featurelab --help

---

Project Structure

---

Author

Shekhar Suman
[e-mail]

---

License

MIT License

---

Keywords

feature-engineering, data-preprocessing, machine-learning, pandas, visualization

---