XPER.viz package

XPER.viz.Visualisation module

XPER.viz.Visualisation. bar_plot(XPER_values, X_test, labels, p, percentage=True)

Create a bar plot to visualize contributions, such as feature importances or data distributions.

Parameters

XPER_valuesnumpy.ndarray

Array of contributions to be visualized.

X_testpandas.DataFrame

Data used for labeling the plot. Typically includes feature names.

labelslist

List of labels corresponding to each contribution in XPER_values.

pint

Number of top contributions to display in the plot.

percentagebool, optional

If True, contributions are shown as percentages. Default is True.

Returns

bar_plot

Examples

from XPER.viz.Visualisation import visualizationClass as viz
                          
labels = list(loan.drop(columns='Loan_Status').columns)
viz.bar_plot(XPER_values=XPER_values, X_test=X_test, labels=labels, p=5, percentage=True)
                          

Example output:

Notes

This function creates a horizontal bar plot using Plotly. The bars represent the contributions, which can be either in absolute values or percentages, depending on the percentage parameter. The function uses a gradient color scheme for the bars, providing a visual distinction of the contribution magnitudes.

XPER.viz.Visualisation. beeswarn_plot(XPER_values, X_test, labels)

Create a beeswarm plot to visualize the distribution of data points across different categories.

Parameters

XPER_valuesnumpy.ndarray

Array of contributions to be visualized in the beeswarm plot.

X_testpandas.DataFrame

Data used for labeling the plot, typically feature names.

labelslist

List of labels corresponding to each contribution in XPER_values.

Returns

beeswarn_plot

Examples

from XPER.viz.Visualisation import visualizationClass as viz
                          
labels = list(loan.drop(columns='Loan_Status').columns)
viz.beeswarn_plot(XPER_values=XPER_values, X_test=X_test, labels=labels)
                          

Example output:

Notes

This function creates a beeswarm plot using Plotly. The plot helps visualize the distribution and density of data points across categories, often used for understanding the spread and concentration of data.

XPER.viz.Visualisation. force_plot(XPER_values, instance, X_test, variable_name, figsize=(16,4))

Create a force plot to explain individual predictions.

Parameters

XPER_valuesnumpy.ndarray

Array of contributions to be visualized.

instanceint

Index of the instance for which the force plot is created.

X_testpandas.DataFrame

Data used for labeling the plot, typically feature names.

variable_namelist

List of variable names corresponding to each contribution in XPER_values.

figsizetuple, optional

Figure size for the plot, default is (16,4).

Returns

force_plot

Examples

from XPER.viz.Visualisation import visualizationClass as viz
                          
labels = list(loan.drop(columns='Loan_Status').columns)
viz.force_plot(XPER_values=XPER_values, instance=1, X_test=X_test, variable_name=labels, figsize=(16, 4))
                          

Example output:

Notes

The force plot visually explains the contributions of features to a specific prediction. It's a useful tool for understanding the influence of different variables on an individual instance's outcome.