src.superphot_plus.plotting.sampling_results
This module contains scripts to plot sampling results.
Module Contents
Functions
|
Plot combined corner plot of all training set samples, excluding |
|
Plot histogram for a posterior parameter. |
|
Plot trace of all posterior samples. |
|
Compare plots of various oversampling methods. |
|
Save all 1d oversampled histograms for each parameter, in one plot. |
|
Plot 2D scatterplots for each pair |
|
Plot the parameter distributions to get better priors for fitting. |
|
Plot 2D UMAP of sampling features. |
|
Plot 2D PACMAP of sampling features. |
Attributes
- plot_corner_plot_all(names, labels, fits_dir, save_dir, aux_bands=Survey.ZTF().priors.aux_bands)[source]
Plot combined corner plot of all training set samples, excluding the overall scaling A.
- Parameters:
names (array-like) – List of object names.
labels (array-like) – Class labels associated the objects in names.
fits_dir (str) – Where object model fits are stored.
save_dir (str) – Path to save the combined corner plot.
aux_bands (array-like, optional) – The auxiliary bands of the fits (for plotting lables). Defaults to ZTF’s aux bands.
- plot_posterior_hist_numpyro_dict(posterior_samples, parameter, output_dir=None)[source]
Plot histogram for a posterior parameter.
- Parameters:
posterior_samples – Dictionary of posterior samples.
parameter – Posterior parameter for which to plot histogram.
output_dir – The directory where to store the plot.
- plot_sampling_trace_numpyro(posterior_samples, output_dir=None)[source]
Plot trace of all posterior samples.
- Parameters:
posterior_samples – The lightcurve samples given by MCMC.
output_dir – The directory where to store the plot.
- compare_oversampling(names, labels, fits_dir, save_dir, allowed_types=SnClass.all_classes(), aux_bands=Survey.ZTF().priors.aux_bands, sampler=None, goal_per_class=1000)[source]
Compare plots of various oversampling methods.
- Parameters:
input_csv (str) – Where supernova list is stored.
allowed_types (array-like) – Types to include in plot.
- plot_oversampling_1d(names, labels, fits_dir, save_dir, priors=Survey.ZTF().priors, sampler='dynesty')[source]
Save all 1d oversampled histograms for each parameter, in one plot. Overlays prior distributions.
- Parameters:
names (array-like) – List of all object names.
labels (array-like) – List of all labels associated with ‘names’.
fits_dir (str) – Directory where model fits are stored.
save_dir (str) – Where to save figure.
priors (MultibandPriors, optional) – Prior object to overlay prior distributions. Defaults to ZTF’s priors.
- plot_combined_posterior_space(names, labels, fits_dir, save_dir, aux_bands=Survey.ZTF().priors.aux_bands)[source]
Plot 2D scatterplots for each pair of fit parameters, to identify clustering among different subclasses.
TODO: modify to plot different points for each type.
- Parameters:
fits_fn (str) – File path for fit posteriors.
save_dir (str) – Where to save figure
- plot_param_distributions(names, labels, fit_folder, save_dir, overlay_gaussians=True, aux_bands=Survey.ZTF().priors.aux_bands)[source]
Plot the parameter distributions to get better priors for fitting.
- Parameters:
fit_folder (str) – Where the posterior fits are stored.
save_dir (str) – Where to save the output figures.
overlay_gaussians (boolean, optional) – Whether to overlay Gaussian estimate of distribution. Defaults to True.
- plot_feature_umap(psg, save_path)[source]
Plot 2D UMAP of sampling features.
- Parameters:
psg (PosteriorSamplesGroup) – The group of posteriors to map
save_path (str) – Where to save the resulting figure.
- plot_feature_pacmap(psg, save_path)[source]
Plot 2D PACMAP of sampling features.
- Parameters:
psg (PosteriorSamplesGroup) – The group of posteriors to map
save_path (str) – Where to save the resulting figure.