iaf.stats

Statistics functionalities.

iaf.stats.hist_bins(values: ndarray, bin_size: float) → tuple[source]

Return the bins to be used for the passed values and bin size.

Parameters:

  • values (np.ndarray) – One-dimensional array of values for which to determine the ideal histogram bins.

  • bin_size (float) – Bin size to use.

Returns:

  • bin_edges (np.ndarray) – Array of bin edges (to use with np.histogram()).

  • bin_centers (np.ndarray) – Array of bin centers.

  • bin_width – Bin width.

iaf.stats.ideal_hist_bins(values: ndarray, scott: bool = False)[source]

Calculate the ideal histogram bins using the Freedman-Diaconis rule.

See: https://en.wikipedia.org/wiki/Freedman%E2%80%93Diaconis_rule

Parameters:

  • values (np.ndarray) – One-dimensional array of values for which to determine the ideal histogram bins.

  • scott (bool) – Whether to use Scott’s normal reference rule (if the data is normally distributed).

Returns:

  • bin_edges (np.ndarray) – Array of bin edges (to use with np.histogram()).

  • bin_centers (np.ndarray) – Array of bin centers.

  • bin_size – Bin width.

iaf.stats.prepare_histogram(values: ndarray, normalize: bool = True, auto_bins: bool = True, scott: bool = False, bin_size: float = 0.0)[source]

Return histogram counts and bins for given values with provided or automatically calculated bin number.

Parameters:

  • values (np.ndarray) – Array of values. It may contain NaNs.

  • normalize (bool) – Whether to normalize the histogram to a probability mass function (PMF). The integral of the PMF is 1.0.

  • auto_bins (bool) – Whether to automatically calculate the bin size from the data.

  • scott (bool) – Whether to use Scott’s normal reference rule (the data should be normally distributed). This is used only if auto_bins is True.

  • bin_size (float) – Bin size to use if auto_bins is False. It will be ignored if auto_bins is True.

Returns:

  • n (np.ndarray) – Histogram counts (optionally normalized to sum to 1.0).

  • bin_edges (np.ndarray) – Array of bin edges (to use with np.histogram()).

  • bin_centers (np.ndarray) – Array of bin centers.

  • bin_width – Bin width.