vadose.rank: Ranking and model performance assessment of vadose model...

In gowusu/vadose: The estimation of saturated hydraulic conductivity and soil water retention Curves in the vadose zone

Description

This function ranks and boxplots BEST, OFEST or lass3 goodness of fit objects. Various models in BEST and OFEST are ranked by either median or mean or max or any of the fun objects.

Usage

vadose.rank(data, var = c("r2", "d", "E", "RMSE", "ARE", "MBE"),
  model = "model", fun = "median", xlab = NULL, ylab = c("R-square",
  "Index of Agreement", "Nash-Sutcliffe", "RMSE", "Average Relative Error",
  "Mean Absolute Error"), main = c("R-square", "Index of Agreement",
  "Nash-Sutcliffe", "RMSE", "Average Relative Error", "Mean Bias Error"),
  ylim = list(c(0.99, 1), c(0.98, 1), c(0.9, 1), c(0, 2.5), c(0, 30), c(-0.2,
  0.2)), las = 2, opar = par(mfrow = c(2, 3), mar = c(7, 5, 2, 1)),
  myrank = NULL)

Arguments

data	dataframe of goodness of fit indicators
var	list of the indicators or variables that should be used to rank the models
model	character. The column name indicating the models that are to be ranked.
fun	Either median or mean or max or min or any of the fun objects
xlab	The x label of the boxplot
ylab	The y label of the boxplot
main	The title of the boxplot
ylim	The y limits of the boxplot
las	The orientation of x labels
opar	The par attributes
myrank	a redundant variable for internal use

Value

rank and fun.summary indicating the aggregation of the data by fun

Author(s)

George Owusu

Examples

data(offinbest) #infiltration data
data(offinpsd) #Particle size Distribution data
data(htheta) #measured h and theta
PSD=offinpsd
Figure2=group.BEST(data = offinbest, group="TownName",PSD=PSD,layout=c(3,4),hlog=TRUE,plot=FALSE)
Figure3=group.OFEST(data = offinbest, hg="best",group="TownName",PSD=PSD,layout=c(3,4),hlog=TRUE,type = "nonlinear",plot=FALSE)
data=rbind(Figure2$statistics2,Figure3$statistics2)
ranking=vadose.rank(data)
ranking$rank

MBE=vadose.rank(data,var="MBE",ylim=c(-1,2),ylab=NULL,main=NULL)
MPE=ranking=vadose.rank(data,var="MPE",ylim=c(-20,30),ylab=NULL,main=NULL)

gowusu/vadose documentation built on May 17, 2019, 7:59 a.m.