This module provides functions for calculating Continuous Ranked Probability Scores. CRPS is a measure of how well a forecast distribution predicts the true values.
You can evaluate forecast performance against future observations by:
Simulating observations for each particle, using the
Loading the simulated observations with
Loading the future observations (once available) with the observation model’s
Calculating the CRPS scores with
- pypfilt.crps.simulated_obs_crps(true_obs, sim_obs)¶
Calculate CRPS scores for simulated observations, such as those recorded by the
SimulatedObstable, against observed values.
The returned array has fields:
true_obs – The table of recorded observations; this must contain the fields
sim_obs – The table of simulated observations; this must contain the fields
ValueError – if
sim_obsdo not contain all of the required fields.
- pypfilt.crps.crps_edf_scalar(true_value, sample_values)¶
Calculate the CRPS for samples drawn from a predictive distribution for a single value, using the probability weighted moment CRPS estimator.
true_value – The (scalar) value that was observed.
sample_values – Samples from the predictive distribution (a 1-D array).
See equation (ePWM) in Zamo and Naveau, 2020.
- pypfilt.crps.crps_sample(true_values, samples_table)¶
Calculate the CRPS score for a table of samples drom from predictive distributions for multiple values, using the empirical distribution function defined by the provided samples.
true_values – A 1-D array of observed values.
samples_table – A 2-D array of samples, where each row contains the samples for the corresponding value in