Outlier_Exterminator is a Python-based tool designed to handle and process outliers in datasets. It provides implementations of the Interquartile Range (IQR), Z-Score, and Percentile methods to detect and treat outliers. The tool supports various outlier treatment techniques, including trimming, capping, and Winsorization.
This project is developed and demonstrated in a Jupyter Notebook for ease of use and understanding.
- Outlier Detection Methods :-
- IQR Method: Detects outliers using the Interquartile Range.
- Z-Score Method: Identifies outliers based on the standard deviation.
- Percentile Method: Flags outliers based on specified percentile thresholds.
- Outlier Treatment Techniques :-
- Trimming: Removes detected outliers from the dataset.
- Capping: Replaces outliers with the nearest valid value within bounds.
- Winsorization: Replaces extreme values with predefined percentiles.
- Customizable Parameters :-
- Configure thresholds for Z-scores.
- Set percentile ranges for trimming and Winsorization.
- Define the IQR multiplier for flexible outlier detection.
- Visualization :-
- Generate boxplots and histograms to visualize the effects of outlier treatment.
- pandas
- numpy
- matplotlib
- seaborn