Skip to contents

Statistical performance indicators

statistical-performance-indicators.Rd

These functions are provided for convenient calculation of modelling the listed model quality objectives performance indicators. All functions work on numeric R vectors, containing modelled or measured concentrations at a single sampling point, or averaged modelled or measured concentrations across multiple sampling points.

  • Bias (vec_bias())

  • Root Mean Square Error (vec_rmse())

  • Maximum Accepted Measurement Uncertainty (scalar_uncer())

  • Root Mean Square Uncertainty (vec_rmsu())

  • Pearson's Correlation coefficient (vec_cor())

  • Standard Deviation (vec_sd())

  • Centred Root Mean Square Error (vec_crmse())

  • MQI (vec_mqi())

Usage

vec_bias(obs, mod, na.rm = FALSE)

vec_rmse(obs, mod, na.rm = FALSE)

scalar_uncer(obs, term, params = mqo_params(), na.rm = FALSE)

vec_rmsu(obs, term, params = mqo_params(), na.rm = TRUE)

vec_cor(obs, mod, na.rm = FALSE)

vec_sd(x, term, na.rm = FALSE)

vec_crmse(obs, mod, na.rm = FALSE)

vec_mqi(obs, mod, term, params = mqo_params(), na.rm = FALSE)

Arguments

obs, mod, x

A numeric vector of observed (measured) concentrations (obs), modelled concentrations (mod), or either of the previous depending on desired output (x).

na.rm

a logical evaluating to TRUE or FALSE indicating whether NA values should be stripped before the computation proceeds.

term

Either "short" or "long", identifying whether data represents Short- or Long-term data.

params

A set of mqor parameters, most simply constructed using mqo_params(). See mqo_params() for more information about how parameter sets can be constructed.

Value

All vec_ functions return a single numeric value. scalar_ functions return a numeric vector the same length as obs or mod.

See also

Other model performance indicators: summarise_mqo_stats(), vec_pi_bias()