FeatureEng-arun8nov 0.0.1

A collection of feature engineering methods for preprocessing datasets.

Feature Engineering Utilities

This Python module provides a collection of feature engineering methods for preprocessing datasets using Pandas. It includes various encoding techniques and scaling methods, which can be used to transform categorical and numerical data for machine learning models.

📦 Class: FeatureEngineering

🔧 Methods

1. OneHot_En(df)

Performs One-Hot Encoding on all columns of the given DataFrame.

• Input: DataFrame with categorical columns.
• Output: DataFrame with one-hot encoded columns.

2. Lable_En(df)

Applies Label Encoding to all columns.

• Input: DataFrame with categorical data.
• Output: DataFrame with integer-encoded values.

3. Order_En(o_list, df)

Applies Ordinal Encoding based on the given list of ordered values.

• Input:
  • o_list: List of lists with ordered values.
  • df: DataFrame with categorical data.
• Output: Ordinal encoded DataFrame.

4. Target_En(df, Target_column_df)

Applies Target Encoding based on the average of the target column for each category.

• Input:
  • df: DataFrame with categorical features.
  • Target_column_df: DataFrame with the target column.
• Output: Encoded DataFrame using target mean values.

5. Freq_En(df)

Applies Frequency Encoding to all columns.

• Input: DataFrame with categorical values.
• Output: DataFrame with frequency-based values.

6. Bin_En(df)

Encodes numeric values as binary string representations based on their position.

• Input: DataFrame with numerical or categorical data.
• Output: DataFrame with binary representations.

7. Stannd_Scale(df)

Applies Standard Scaling (Z-score) to numerical columns.

• Input: DataFrame with numerical features.
• Output: DataFrame with standardized values.

8. Normm_Scale(df)

Applies Min-Max Normalization to numerical columns.

• Input: DataFrame with numerical features.
• Output: DataFrame with normalized values between 0 and 1.

---

🚀 How to Use

from Preprocess import  FeaturEngineering
import pandas as pd

df = pd.DataFrame({...})  # your input data
fe = Preprocess()

encoded_df = fe.OneHot_En(df)       # One-Hot Encoding
labeled_df = fe.Lable_En(df.copy()) # Label Encoding