gofFun {spsh}

Goodness-of-fit and Information Criteria

Description

Calculates goodness-of-fit criteria and the likelihood-based Akaike and Bayesian Information Criterion based on a given parameter set, typically from the estimation procedure.

Usage

gofFun(
  phat,
  shpmodel = "01110",
  retdata = NULL,
  condata = NULL,
  weight,
  psel,
  ivap.query = NULL,
  hclip.query = FALSE
)

Arguments

phat

Vector of non-transformed (back-transformed) model parameters after estimation, i.e. the best fit or maximum likelihood estimate.

shpmodel

Character specifying the soil hydraulic property model.

retdata

Dataframe or matrix with 2 columns. The first with pressure head values in log₁₀ [cm], i.e. pF values, and the second with volumetric water contents in [cm cm-3].

condata

Dataframe or matrix with 2 columns. The first with pressure head values in log₁₀ [cm], i.e. pF values, and the second with hydraulic conductivity values log₁₀ [cm d-1].

weight

List of the model residual weights used or estimated by the parameter estimation scheme, to calculate the weighted statistical analyses.

psel

Vector specifying the selected parameters for the parameter estimation from parL.

ivap.query

Specification of ivap method, if FALSE selected, no isothermal vapour conductivity is consideredSee Details

hclip.query

Implemented purely for future compatability. Currently no use. See Details

Details

Output for data groups.

th	list with goodness of fit statistics for the retention curve see below
logKh	listwith output same as th but for the log10 fitted conductivity curve
combined	list with AIC, AICc, and BIC calculated for the multi-objective function if arguments retdata and condata are both !NULL

Statistical analyses of the inverse modelling results.

me	mean (weighted) error
mae	mean absolute (weighted) error
mse	mean squared (weighted) error
rss	sum of squared (weighted) errors
rmse	root mean squared (weighted) error
AIC	Akaike Information Criteria
AICc	corrected Akaike Information Criteria
BIC	Bayesian Information Criteria
m	number of observations

Author(s)

Tobias KD Weber , [e-mail]

References

Höge M, Wöhling T, Nowak W (2018). “A Primer for Model Selection: The Decisive Role of Model Complexity.” Water Resources Research, 54(3), 1688–1715. doi: 10.1002/2017WR021902.

Examples

data("shpdata1")
retdata <- shpdata1$TS1$wrc
condata <- shpdata1$TS1$hcc
condata <- condata[!is.na(condata[,1]),]
# Parameter list
parL <- list("p" = c("thr"= 0.05, "ths" = 0.45, "alf1" = 0.01, "n" = 2, "Ks" = 100, "tau" = .5),
             "psel" = c(1, 1, 0, 1, 1, 1),
             "plo" = c(0.001 , 0.2, 0.001, 1.1, 1, -2),
             "pup" = c(0.3, 0.95, 1, 10, 1e4, 10)
             )
# Calulation of the goodness of fit.
gofL <-gofFun(parL$p, shpmodel = "01110", retdata = retdata, condata = condata, 
              weight = weightFun(weightmethod = "fix1"), parL$psel, 
              ivap.query = NULL, hclip.query = FALSE)