Parameter sensitivity {femmeR} | R Documentation |
Using Application=Parcovariance in FEMME a .PCV file is generated which can be read and further processed with these functions
read.pcv(pcvfile) ## sensitivity analysis ## S3 method for class 'pcv': summary(object, parnames = NULL,order=TRUE,...) ## S3 method for class 'pcv': plot(x,xvar=seq(along=sens.data[,1]),ylim=NULL,pari=seq(1,length(x$sens)-4),parnames=NULL,type='b',scale=FALSE,...) ## S3 method for class 'pcv': points(x,xvar=seq(along=sens.data[,1]),pari=1,type='b',...) ## Collinearity tables collin.format.table(x, parnames = NULL) collin.remove(x, pari)
pcvfile |
File ending in .pcv generated by FEMME |
x |
object of class pcv created by read.pcv |
object |
object of class pcv created by read.pcv |
ylim |
Limits of the Y axis |
parnames |
Parameter names |
order |
Order parameters in decreasing importance |
xvar |
Variable such as time or depth |
pari |
Vector of parameter index |
scale |
A logical value indicating if Y axis should be common to all parameters |
type |
Plot type |
... |
Additional plot parameters |
plot.pcv
plots by default the sensitivity functions of all
parameters. You can also specify to plot only a single parameter.
Adding a second set of sensitivities can be done with points.pcv
read.pcv
returns a list with components
sens |
Relative sensitivity of parameters at observed data points |
collin |
Collinearity index for all possible parameter combinations |
...
Use at your own risk
Henrik Andersson <h.andersson@nioo.knaw.nl>
Soetaert et al, 2002, Ecological Modelling 151: 177-193, Brun et al, 2001, Water Resources Research 37: 1015-1030
## For examples see: vignette("femmeR")