✎ᝰ plotting.vif¢

The VIFPlotter class provides functionality to visualize the Variance Inflation Factor (VIF) for features in a dataset. This module helps users detect multicollinearity by generating informative bar plots of VIF values for each feature.

OverviewΒΆ

The module exposes a plotter class that:

  • Accepts a pandas DataFrame with VIF values and feature names

  • Plots VIF values for each feature as a horizontal bar chart

  • Highlights common VIF thresholds (5 and 10) for interpretation

  • Supports output directory customization and plot styling

Class ReferenceΒΆ

class owlmix.plotting.vif.VIFPlotParamsΒΆ

Dataclass for specifying VIF plotting parameters.

class owlmix.plotting.vif.VIFPlotter(data, params)ΒΆ

Plots the Variance Inflation Factor (VIF) for features in a pandas DataFrame.

Parameters:

  • data (pandas.DataFrame) – Input DataFrame containing feature names, VIF values, and colors.

  • params (VIFPlotParams) – Configuration parameters for VIF plotting.

generate(output_dir: str = 'outputs/charts') → strΒΆ

Generates and saves a horizontal bar plot of VIF values for each feature.

Parameters:

output_dir (str) – Directory to save the generated plot.

Returns:

File path to the saved VIF chart image.

Return type:

str

Sample OutputΒΆ

The VIF plot consists of a horizontal bar chart where each bar represents the VIF value for a feature. Dashed vertical lines at VIF=5 and VIF=10 indicate common thresholds for multicollinearity concerns. VIF values are annotated on each bar for clarity.

Sample VIF Plot

ReferencesΒΆ

Back to Home