A set of functions that can be called on an cfp_pfres object (the result of a call to pro_flux) to assess the quality of the model.
Usage
error_concentration(x, param_cols = NULL, normer = "sd")
# S3 method for class 'cfp_pfres'
error_concentration(x, param_cols = NULL, normer = "sd")
# S3 method for class 'cfp_fgres'
error_concentration(x, param_cols = NULL, normer = "sd")
# S3 method for class 'cfp_altres'
error_concentration(x, param_cols = NULL, normer = "sd")
error_efflux(x, param_cols, EFFLUX, normer = "sd", ...)
# S3 method for class 'cfp_pfres'
error_efflux(x, param_cols, EFFLUX, normer = "sd", ...)
# S3 method for class 'cfp_fgres'
error_efflux(x, param_cols, EFFLUX, normer = "sd", ...)
# S3 method for class 'cfp_altres'
error_efflux(x, param_cols, EFFLUX, normer = "sd", ...)Arguments
- x
A cfp_pfres object, that is returned by a call to pro_flux()
- param_cols
The columns that, together, define different parameters (e.g. different gases) for which NRMSEs should be calculated separately (e.g. "gas"). Defaults to the id_cols of layers_map. If no such distinction is wished, set to
character()- normer
a character string defining the type of normalization to be applied. Can be one of
- mean
the arithmetic mean of a
- sd
the standard deviation of a (default).
- range
the difference between the range of a
- IQR
the difference between the interquantile range of a
- EFFLUX
A data.frame with (at most) one value of efflux per profile of x. Must contain any id_cols of x needed.
- ...
Further arguments passed to efflux
Value
The calculated error estimate for a single model, a list of models
(cfp_altres) and for each parameter combination in param_cols
Details
For error_concentration, the way the error parameter is calculated for cfp_fgres and cfp_pfres objects is entirely different and should not be used in comparison between the two. NRMSE of cfp_pfres objects are calculated as the mean of depth-wise NRMSEs of modelled versus input gas concentrations. 'NRMSE's of cfp_fgres objects simply calculate the mean of (dcdz_sd / dcdz_ppm) per group described in param_cols.
Examples
PROFLUX <- pro_flux(base_dat)
error_concentration(PROFLUX)
#> # A tibble: 2 × 3
#> # Groups: site [2]
#> site gas NRMSE
#> <chr> <chr> <dbl>
#> 1 site_a CO2 0.209
#> 2 site_b CO2 0.237
error_efflux(
PROFLUX,
EFFLUX = data.frame(efflux = 1),
param_cols = c("site"))
#> # A tibble: 2 × 2
#> site NRMSE
#> <chr> <dbl>
#> 1 site_a 1.45
#> 2 site_b 1.39