Summary statistics of the model probabilities on an object of class hmsc

Share:

Description

The function quantile calculates quantiles on the probabilities of a model constructed with hmscH. Quantiles are calculated for each species included in the model. The function basicStatHMSC calculates the means and standard deviations on the probabilities of a model constructed with hmscH. The function summary calculates the means, standard deviations, and three quantiles (0.025,0.5,0.975) on the probabilities of a model constructed with hmscH.

Usage

1
2
3
4
5
## S3 method for class 'hmsc'
summary(object, ...)
## S3 method for class 'hmsc'
quantile(x, probs = c(0.025, 0.5, 0.975), type = 8, ...)
basicStatHMSC(x, stats = c("mean", "sd"))

Arguments

x

An object of class hmscH An object of class hmsc

object

An object of class hmscH

probs

A numeric vector of probabilities with values ranging from 0 to 1. See quantile for more details.

type

an integer between 1 and 9 selecting one of the nine quantile algorithms detailed in quantile. Default is 8 as suggested by Hyndman and Fan (1996).

stats

A character vector defining the statistics to calculate. Any unambiguous variation of wording used in this argument is accepted.

...

Arguments passed to quantile or to summary.

Details

These functions can calculate summary statistics for any type of hmscH object, whether they stem from a linear or a non-linear model.

Value

For quantile: A three-dimensional array where the first dimension represents sites, the second dimension represents species, and the third dimensions the quantiles. For basicStatHMSC: A matrix where the rows represents sites and the columns represents species if only the mean or the standard deviation is calculated. Otherwise, a three-dimensional array where the first dimension represents sites, the second dimension represents species, and the third dimensions the means and standard deviations.

Author(s)

F. Guillaume Blanchet

References

Hyndman, R. J. and Fan, Y. (1996) Sample quantiles in statistical packages, American Statistician, 50, 361-365.

See Also

quantile, hmscH

Examples

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
#==================
### Simulating data
#==================
nspecies<-20
desc<-cbind(1,scale(1:50),scale(1:50)^2)
dataBase<-communitySimulH(desc,nsp=nspecies)

#=================================
### Formatting data and parameters
#=================================
formdata<-as.HMSCdata(dataBase$data$Y,desc,Ypattern="sp",interceptX=FALSE,interceptTr=FALSE)

#========================
### Formatting parameters
#========================
startParamX<-matrix(rnorm(nspecies*ncol(desc)),nrow=nspecies,ncol=ncol(desc))
startMean<-rnorm(ncol(desc))
startSigma<-rWishart(1,ncol(desc)+1,diag(ncol(desc)))[,,1]

formparam<-as.HMSCparam(formdata,paramX=startParamX,means=startMean,sigma=startSigma)
formpriors<-as.HMSCprior(formparam,rep(0,length(formparam$means)),0,length(formparam$means),diag(length(formparam$means)))

#===================================================
### Building model using a linear modelling approach
#===================================================
modelLinear<-hmscH(formdata,formparam,formpriors,niter=200,nburn=100)

meansdHMSC<-basicStatHMSC(modelLinear)
quantileHMSC<-quantile(modelLinear)
summaryHMSC<-summary(modelLinear)