jesterTOV.inference.postprocessing module#
Visualization and analysis tools for EOS inference results.
After a sampling run completes, the postprocessing module produces a standard suite of diagnostic
and publication-quality plots from the HDF5 result file written by
InferenceResult. The CLI entry point
run_jester_postprocessing (see main()) can be called with the same config.yaml
used for inference.
Configuration#
Postprocessing behaviour is controlled through the postprocessing block of the inference
configuration file, parsed into PostprocessingConfig.
Data Loading#
|
Load EOS data from the specified output directory. |
|
Load prior EOS data for comparison. |
|
Load injection EOS data from an NPZ file. |
Plotting Functions#
The main entry point is generate_all_plots(), which calls each individual plot function in
sequence. Individual functions can also be called directly for custom workflows.
|
Generate selected plots for the specified output directory. |
|
Create a cornerplot for EOS parameters. |
|
Create a mass-radius plot with posterior probability colouring. |
|
Create a mass-Lambda plot with posterior probability colouring. |
|
Create a mass-Lambda ratio plot relative to the injection. |
|
Create an equation-of-state plot (pressure vs density). |
|
Create a speed-of-sound-squared vs density plot. |
|
Create KDE histograms for key EOS-derived parameters. |
|
Create a contour plot of radii vs mass. |
|
Create a contour plot of pressure vs density. |
Utilities#
|
Configure matplotlib with TeX rendering and sensible defaults. |
|
Compute a symmetric credible interval around the median. |
|
Run postprocessing from a YAML config file. |