opendvp.tl.spatial_hyperparameter_search

opendvp.tl.spatial_hyperparameter_search(adata, x_y=None, threshold_range=None, return_df=False, plot_network_at=None)

Perform a hyperparameter search over a range of threshold values.

To determine the number of connected nodes and average neighbors for different threshold values, and optionally plot the network of connected nodes at a given threshold.

Return type:

DataFrame | tuple

Parameters

adataAnnData

Spatially indexed data.

x_ylist of str, default [‘x_centroid’, ‘y_centroid’]

Column names in adata.obs representing the spatial coordinates.

threshold_rangenp.ndarray, default np.arange(1, 100, 1)

Range of threshold values to test.

return_dfbool, default False

If True, return the DataFrame with threshold statistics along with the plot.

plot_network_atOptional[int], default None

The threshold value at which to plot the network of connected nodes. If None, no plot is generated.

Returns:

If return_df is True:

tuple[pandas.DataFrame, tuple[matplotlib.figure.Figure, matplotlib.axes.Axes]]

DataFrame with threshold statistics and the plot (figure, axes).

Else:

tuple[matplotlib.figure.Figure, matplotlib.axes.Axes]

The plot (figure, axes).