eff_par: Compute direct, indirect and total effects (or impacts) for...

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

Compute direct, indirect and total effects (or impacts) for continous parametric covariates in PS-SAR regression models.

Description

Compute direct, indirect and total effects (also named impacts) for non-parametric covariates included in a semiparametric spatial or spatio-temporal SAR model. This model must include a spatial lag of the dependent variable (SAR) to have indirect effects different from 0, otherwise, total and direct effects are the same.

Usage

eff_par(sptsarfit, variables, nrep = 1000, seed = 1111, m = 100,
  p = 50, tol = 0.01)

Arguments

sptsarfit	A psar object fitted using psar function
variables	vector including names of non-parametric covariates.
nrep	number of repetitions for the simulation. Default 1000.
seed	initial seed to get random numbers. Must be set to a specific value to make reproducible results. Default 1111.
m	number of powers to compute a vector of traces of powers of a spatial weight matrix (see trW in spdep package). Default 100.
p	number of samples used in MC simulation of traces of a spatial weight matrix (see trW in spdep package).Default 50.
tol	tolerance (relative to largest variance) for numerical lack of positive-definiteness in Sigma when simulate βs from the maximum likelihood estimates (see mvrnorm in MASS package). Default 0.01.

Details

DESCRIBE ALGORITHM TO SIMULATE PARAMETRIC EFFECTS

Value

An object of class par.eff.psar. Can be printed with summary.

The object returned is a list with 3 matrices including the results of simulated effects:

tot_eff	Matrix including simulated total effects for each variable in rows.
dir_eff	Matrix including simulated direct effects for each variable in rows.
ind_eff	Matrix including simulated indirect effects for each variable in rows.

References

• LeSage, J. and Pace, K. (2009). Introduction to Spatial Econometrics. CRC Press, Boca Raton.

See Also

• psar estimate spatial or spatio-temporal semiparametric PS-SAR regression models.

• eff_nopar compute total, direct and indirect effect functions for non-parametric continuous covariates.

• fit_terms compute smooth functions for non-parametric continuous covariates.

• impacts similar function in spdep package to compute impacts in spatial parametric econometric models.

Other Direct, Indirect and Total Effects.: eff_nopar, plot_eff_nopar

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

######################  No Spatial Trend: PSAR including a spatial 
######################  lag of the dependent variable
form1 <- unrate ~ partrate + agri + cons +
                 pspl(serv,nknots=15) +
                 pspl(empgrowth,nknots=20) 
 gamsar <- psar(form1,data=unemp_it,sar=TRUE,Wsp=Wsp_it)
 summary(gamsar)
 ###### Parametric Total, Direct and Indirect Effects
 list_varpar <- c("partrate","agri","cons")
 eff_parvar <- eff_par(gamsar,list_varpar)
 summary(eff_parvar)
 
######################   PSAR-ANOVA with spatial trend
form2 <- unrate ~ partrate + agri + cons +
                  pspl(serv,nknots=15) + pspl(empgrowth,nknots=20) +
                  pspt(long,lat,nknots=c(20,20),psanova=TRUE,
                  nest_sp1=c(1,2),nest_sp2=c(1,2))
##### Spatial trend fixed for period 1996-2014
geospanova_sar <- psar(form2,data=unemp_it,Wsp=Wsp_it,sar=TRUE,
                   control=list(thr=1e-1,maxit=200,trace=FALSE))
summary(geospanova_sar)
 ###### Parametric Total, Direct and Indirect Effects
 list_varpar <- c("partrate","agri","cons")
 eff_parvar <- eff_par(geospanova_sar,list_varpar)
 summary(eff_parvar)
 
######################   PSAR-ANOVA with spatio-temporal trend and 
######################   temporal autorregresive noise
 form3 <- unrate ~ partrate + agri + cons +
                   pspl(serv,nknots=15) + pspl(empgrowth,nknots=20) +
                   pspt(long,lat,year,nknots=c(18,18,8),psanova=TRUE,
                   nest_sp1=c(1,2,3),nest_sp2=c(1,2,3),
                   nest_time=c(1,2,2),ntime=19)
sptanova_sar_ar1 <- psar(form3,data=unemp_it,Wsp=Wsp_it,sar=TRUE,ar1=TRUE,
                    control=list(thr=1e-1,maxit=200,trace=FALSE))
summary(sptanova_sar_ar1)
 ###### Parametric Total, Direct and Indirect Effects
 list_varpar <- c("partrate","agri","cons")
 eff_parvar <- eff_par(sptanova_sar_ar1,list_varpar)
 summary(eff_parvar)

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