anova.psar: Build an ANOVA table of fitted semiparametric models.

In rominsal/sptpsar: Spatio-Temporal Semiparametric Models with Spatial Lags

Build an ANOVA table of fitted semiparametric models.

Description

Method to compare goodness-of-fit measures of fitted spatio-temporal semiparametric PS-SAR regression models.

Usage

## S3 method for class 'psar'
anova(object, ...)

Arguments

object	a fitted semiparametric model of class psar.
...	additional fitted models to be compared.

Value

A class anova.psar table including goodness-of-fit measures for each fitted model. This table is also printed.

Author(s)

Roman Minguez [e-mail]

Examples

################################################
 ###################### Examples using a panel data of rate of
 ###################### unemployment for 103 Italian provinces in period 1996-2014.
library(sptpsar)
data(unemp_it); Wsp <- Wsp_it
 ######################  GAM pure
form1 <- unrate ~ partrate + agri + cons +
                 pspl(serv,nknots=15) +
                 pspl(empgrowth,nknots=20)
gam <- psar(form1,data=unemp_it)
gamsar <- psar(form1,data=unemp_it,sar=TRUE,Wsp=Wsp_it)
form2 <- unrate ~ partrate + agri + cons +
  pspl(serv,nknots=15) + pspl(empgrowth,nknots=20) +
  pspt(long,lat,nknots=c(20,20),psanova=TRUE)
gamsp2d <- psar(form2,data=unemp_it)
gamsp2dsar <- psar(form2,data=unemp_it,sar=TRUE,Wsp=Wsp_it)
anova(gam,gamsar,gamsp2d,gamsp2dsar)

rominsal/sptpsar documentation built on June 1, 2022, 2:03 a.m.