linreg_ally.plotting

Functions

qq_and_residuals_plot(y_actual, y_predicted[, concatenate])

Generate a Q-Q plot (standardized residuals vs theoretical quantiles)

Module Contents

linreg_ally.plotting.qq_and_residuals_plot(y_actual, y_predicted, concatenate=True)[source]

Generate a Q-Q plot (standardized residuals vs theoretical quantiles) and a Residuals vs. Fitted Values plot using Altair for regression diagnostics.

Parameters:

  • y_actual (array-like) – Actual observed values from the dataset. Must be numeric and have the same length as y_predicted.

  • y_predicted (array-like) – Predicted (fitted) values from the regression model. Must be numeric and have the same length as y_actual.

  • concatenate (bool, optional (default=True)) – If True, concatenates the Q-Q plot and Residuals vs. Fitted Values plot side by side. If False, returns the plots separately as individual Altair charts.

Returns:

  • If concatenate=True, returns a concatenated Altair chart with both plots.

  • If concatenate=False, returns a tuple containing the Q-Q plot and Residuals vs. Fitted Values plot.

Return type:

alt.Chart or tuple of alt.Chart

Raises:

  • ValueError – If y_actual or y_predicted are empty, have mismatched lengths, or contain fewer than two data points.

  • TypeError – If y_actual or y_predicted are not numeric.

Examples

>>> import numpy as np
>>> np.random.seed(42)
>>> y_actual = np.random.normal(10, 2, 100)
>>> y_predicted = y_actual + np.random.normal(0, 1, 100)
>>> chart = qq_and_residuals_plot(y_actual, y_predicted, concatenate=True)
>>> chart.display()