ResidualPlot

class gaitmap_challenges.visualization.ResidualPlot(cv_result: DataFrame, prediction_col_name: str, reference_col_name: str, metric_name: Optional[str] = None)

Plot a residual (similar to Blant-Altman) plot to visualize the error dependency.

Most parameters are shared between both plotting backends. Specific parameters are prefixed with bokeh_ or matplotlib_ or are expected to be passed to the method directly.

To use this visualization, you need to make the “raw” predictions available in the cv_result. This is usually done by including the with the NoAgg wrapper in the scorer output.

Parameters:

cv_result

The CV results of a single algorithm as loaded by load_run.

prediction_col_namestr

The name of the column in the cv_results that contains the predictions.

reference_col_namestr

The name of the column in the cv_results that contains the reference values.

metric_namestr, optional

An optional display name for the used metric. If not provided, the column name is used.

Attributes:

metric_name

Methods

bokeh()	Create the plot using bokeh.
matplotlib([ax])	Create the plot using matplotlib.

__init__(cv_result: DataFrame, prediction_col_name: str, reference_col_name: str, metric_name: Optional[str] = None) → None

bokeh()

Create the plot using bokeh.

This creates a plot object that can be displayed using bokeh.plotting.show.

matplotlib(ax=None)

Create the plot using matplotlib.

You can optionally pass an existing matplotlib axes object to plot into.