metrics

The following methods may be used to describe the fit between the model simulation and the observations.

Examples

Calculate metrics directly:

ps.stats.rmse(sim, obs)

Or from a Pastas model:

ml.stats.rmse()

Functions

mae(→ float)	Compute the (weighted) Mean Absolute Error (MAE).
rmse(→ float)	Compute the (weighted) Root Mean Squared Error (RMSE).
sse(→ float)	Compute the Sum of the Squared Errors (SSE).
pearsonr(→ float)	Compute the (weighted) Pearson correlation (r).
evp(→ float)	Compute the (weighted) Explained Variance Percentage (EVP).
nse(→ float)	Compute the (weighted) Nash-Sutcliffe Efficiency (NSE).
nnse(→ float)	Compute the (weighted) Normalized Nash-Sutcliffe Efficiency (NNSE).
rsq(→ float)	Compute R-squared, possibly adjusted for the number of free parameters.
bic(→ float)	Compute the Bayesian Information Criterium (BIC).
aic(→ float)	Compute the Akaike Information Criterium (AIC).
aicc(→ float)	Compute the Akaike Information Criterium with second order
kge(→ float)	Compute the (weighted) Kling-Gupta Efficiency (KGE).
picp(obs, bounds)	Compute the prediction interval coverage probability (PICP).