src.plot.receptive_fields_distribution_plot

Classes

ReceptiveFieldsDistributionPlot(**kwargs)

class src.plot.receptive_fields_distribution_plot.ReceptiveFieldsDistributionPlot(**kwargs)
Author:

Alberto M. Esmoris Pena

Class to plot the distribution of predicted or expected value in the receptive fields. See ReceptiveFieldsDistributionPlot.

Variables:

  • y_rf (np.ndarray) – The expected value for each point for each receptive field.

  • yhat_rf (np.ndarray) – The predicted value for each point for each receptive field.

  • class_names (list) – The names representing each class.

__init__(**kwargs)

Initialize a MplPlot.

Parameters:

kwargs – The attributes for the MplPlot.

plot(**kwargs)

Plot the distribution of predicted and/or expected values.

See plot.Plot.plot().

count(y)

Compute the absolute frequency and the count of receptive fields containing at least one class case.

Parameters:

y (np.ndarray or list) – Either the predicted or the expected values for each receptive field. In other words, y[i] is the i receptive field such that y[i][j] is the value corresponding to the point j in the receptive field i.

Returns:

The absolute frequency considering all points and the count of how many receptive fields do contain at least one element of a given class. If a and b are the returned vectors, a[i] is the absolute frequency of class i and b[i] is the number of receptive fields that contain at least one point of class i.

Return type:

tuple

plot_counts(x, fig, count, rf_count, axes=None, count_title=None, rf_count_title=None, subplot_offset=0, nrows=1, ncols=2)

Do the plots corresponding to a set of counts. There are two potential sets of counts, the one corresponding to predictions, and the one corresponding to expected values.

Parameters:

  • x (np.ndarray) – The values for the x-axis of the plots.

  • fig (mpl.figure.Figure) – The matplotlib figure where the plots belong to.

  • count (np.ndarray) – How many cases among all receptive fields. In other words, count[i] is the number of cases corresponding to class i.

  • rf_count (np.ndarray) – How many receptive fields contain at least one case for each class. In other words, rf_count[i] is the number of receptive fields that contain at least one point of class i.

  • axes (list) – List where the axes of the generated plots will be appended.

  • count_title – The title for the subplot corresponding to count.

  • rf_count_title – The title for the subplot corresponding to rf_count.

  • subplot_offset (int) – How many subplots skip before generating the new ones.

  • nrows (int) – How many rows of subplots.

Type:

count_title: str

Paran ncols:

How many columns of subplots.

Returns:

Nothing at all, but the plots are made effective. If axes is not None, the corresponding axes are appended to the list.