plot_betti_surfaces

gtda.plotting.plot_betti_surfaces(betti_curves, samplings=None, homology_dimensions=None, plotly_params=None)[source]

Plot Betti surfaces (Betti numbers against “time” and filtration parameter) by homology dimension.

Parameters
  • betti_curves (ndarray of shape (n_samples, n_homology_dimensions, n_bins)) – Collection whose each entry contains the Betti numbers for n_homology_dimensions discretised Betti curves. Index i along axis 1 is assumed to correspond to homology dimension i.

  • samplings (ndarray of shape (n_homology_dimensions, n_bins)) – Filtration parameter values to be used as one of the independent variables when plotting the Betti surfaces. The other independent variable is “time”, i.e. the sample index.

  • homology_dimensions (list, tuple or None, optional, default: None) – Homology dimensions for which the Betti surfaces should be plotted. If None, all available dimensions will be used.

  • samplings – For each homology dimension, (filtration parameter) values to be used on the x-axis against the corresponding values in betti_curves on the y-axis.

  • plotly_params (dict or None, optional, default: None) – Custom parameters to configure the plotly figure. Allowed keys are "traces" and "layout", and the corresponding values should be dictionaries containing keyword arguments as would be fed to the update_traces and update_layout methods of plotly.graph_objects.Figure.

Returns

figs/fig – If n_samples > 1, a tuple of figures representing the Betti surfaces, with one figure per dimension in homology_dimensions. Otherwise, a single figure representing the Betti curve of the single sample present.

Return type

tuple of plotly.graph_objects.Figure/ plotly.graph_objects.Figure object