maria jesus herrerias
2021-Apr-14 21:57 UTC
[R] different dimensions in W and my data in moran and spatial model
Hello everyone,
I am writing my code in R to run the spatial models. I am learning this
at the moment.
I have my data for 116 countries over the period 1990-2014,
and I got the shapefile from the website below.
The objective is to get the matrix to run the spatial models.
In the code below, I can get the matrix for the world and also for the
subsample.
But when I try to run the moran test or sar model I get an error.
The error is related to the dimension of W (listw) and my dataset. So
any help,
I would be grateful.
Below my code just in case someone can help here.
# Load the required Packages.
`library(spdep) # Spatial Dependence: Weighting Schemes,
Statistics
`library(rgdal) # to read shapefiles
#shapefiles (world map from diva-gis)
# Read the panel data from stata format for the regression analysis
below.
`library(foreign)
`mydata <- read.dta("C:/Users/Usuario/Desktop/instituciones/revision
energy economics/
spatial panel in stata/countries_shp/neworiginal.dta")
`fix(mydata)
`attach(mydata)
my data contains 116 countries from 1990 to 2014 in panel.
# Prepare the dataset for panel data analysis in R.
`library(plm)
`mydata <- pdata.frame(mydata, index = c("id",
"year"))
# Read the shapefile and set up the working directory.
`setwd("C:/Users/Usuario/Desktop/instituciones/revision energy
economics/spatial panel in stata/countries_shp")
`worlddata <- readOGR("countries.shp")
`names(worlddata)
`summary(worlddata)
# Get the centroids
`coords <-coordinates(worlddata)
A. Contiguity based relations
1. First Order Queen Contiguity
`queen.nb = poly2nb(worlddata, queen=TRUE)
`summary(queen.nb)
`plot(queen.nb,coords)
# the numbers of neighbours of regions in the neighbours list.
`cards <- card(queen.nb)
# convert into a listw type based on contiguity row normalized
`queen.listw <- nb2listw(queen.nb,style="W", zero.policy =
TRUE)
`summary(queen.listw,zero.policy=TRUE)
#marginal effects
`impacts(reg2, listw = listw1,zero.policy=TRUE)
#marginal effects with p-values
`summary(impacts(reg2, listw = listw1), R= 500, zstats
TRUE,zero.policy=TRUE)
`attributes(queen.listw)
#Selection of a subset of countries to fit with mydataset (116
countries, the previous ones I did for all countries in the shapefile)
`setwd("C:/Users/Usuario/Desktop/instituciones/revision energy
economics/spatial panel in stata/countries_shp")
`worlddata <- readOGR("countries.shp")
`queen.nb = poly2nb(worlddata, queen=TRUE)
`coords <-coordinates(worlddata)
`plot(queen.nb, coords)
Following Roger Bivand example:
`to.be.dropped <- c(1, 2, 5, 6, 8, 9, 10, 13, 14, 18, 20, 21, 23, 26,
28, 29, 31, 34, 36, 37, 39, 40, 41, 45, 46, 47, 48, 51, 52, 53, 55, 56, 58,
59, 62, 63, 64, 66, 68, 69, 74, 75, 77, 78, 79, 80, 81, 82, 85, 86, 87, 89,
90, 94, 95, 97, 98, 99, 100, 102, 103, 104, 105, 107, 109, 110, 118, 123,
124, 125, 127, 130, 131, 133, 134, 137, 138, 139, 140, 143, 145, 146, 148,
149, 151, 152, 153, 155, 157, 159, 161, 162, 166, 167, 170, 173, 175, 176,
177, 178, 181, 182, 183, 184, 186, 190, 193, 194, 195, 198, 199, 200, 201,
202, 203, 204, 205, 206, 207, 208, 211, 212, 213, 215, 216, 218, 219, 221,
222, 224, 226, 227, 228, 229, 232, 233, 237, 239, 240, 242, 246, 247, 248,
255, 258, 259, 260, 261, 265)
sub.queen.nb <- subset(queen.nb,
!(1:length(queen.nb) %in% to.be.dropped))
which(!(attr(queen.nb, "region.id") %in%
attr(sub.queen.nb, "region.id")))
`sub.queen.listw <- nb2listw(sub.queen.nb,style="W",
zero.policy = TRUE)
so now, I have my 116 countries in the queen matrix.
# My data and variables
`y <- cbind(lei)
`x <- cbind(lgdppcnew, lgdppcnew2, industryg, importsg, fdig)
`xy <- cbind(mydata$x, mydata$y)
`listw1 <- sub.queen.listw
`coords <- coords
#Define formula
`reg.ols <- y ~ x
# Autocorrelation test
`lm.morantest(ols.eq1,listw1,zero.policy=TRUE)
`plot.moran(ols.eq1,listw1)
(here I get the error where it is said the objects have different
dimensions)
`reg2 <- lagsarlm(ols.eq1, data = mydata, listw1, zero.policy=TRUE)
`summary(reg2,correlation=TRUE, Nagelkerke=TRUE, Hausman=TRUE,
zero.policy=TRUE)
(here I get the error where it is said the objects have different
dimensions)
#marginal effects
`impacts(reg2, listw = listw1,zero.policy=TRUE)
#marginal effects with p-values
`summary(impacts(reg2, listw = listw1), R= 500, zstats
TRUE,zero.policy=TRUE)
I don't know if it is for my panel data, the missing values or
something else.
I tried also in Stata having similar issues.
thanks in advance and my apologies if it is very naive question, but I
am stuck.
all the best
Maria Jesus
[[alternative HTML version deleted]]
Bert Gunter
2021-Apr-14 22:42 UTC
[R] different dimensions in W and my data in moran and spatial model
Almost certainly better posted on R-Sig-geo, not here. See here for more info on R mailing lists: https://www.r-project.org/mail.html Bert Gunter "The trouble with having an open mind is that people keep coming along and sticking things into it." -- Opus (aka Berkeley Breathed in his "Bloom County" comic strip ) On Wed, Apr 14, 2021 at 3:32 PM maria jesus herrerias < mjherreriast at gmail.com> wrote:> Hello everyone, > > I am writing my code in R to run the spatial models. I am learning this > at the moment. > I have my data for 116 countries over the period 1990-2014, > and I got the shapefile from the website below. > The objective is to get the matrix to run the spatial models. > In the code below, I can get the matrix for the world and also for the > subsample. > But when I try to run the moran test or sar model I get an error. > The error is related to the dimension of W (listw) and my dataset. So > any help, > I would be grateful. > > Below my code just in case someone can help here. > > # Load the required Packages. > > `library(spdep) # Spatial Dependence: Weighting Schemes, > Statistics > `library(rgdal) # to read shapefiles > > #shapefiles (world map from diva-gis) > > # Read the panel data from stata format for the regression analysis > below. > > `library(foreign) > > `mydata <- read.dta("C:/Users/Usuario/Desktop/instituciones/revision > energy economics/ > spatial panel in stata/countries_shp/neworiginal.dta") > > `fix(mydata) > `attach(mydata) > > my data contains 116 countries from 1990 to 2014 in panel. > > # Prepare the dataset for panel data analysis in R. > > `library(plm) > > `mydata <- pdata.frame(mydata, index = c("id", "year")) > > > # Read the shapefile and set up the working directory. > > `setwd("C:/Users/Usuario/Desktop/instituciones/revision energy > economics/spatial panel in stata/countries_shp") > > `worlddata <- readOGR("countries.shp") > `names(worlddata) > `summary(worlddata) > > # Get the centroids > > `coords <-coordinates(worlddata) > > A. Contiguity based relations > > 1. First Order Queen Contiguity > > `queen.nb = poly2nb(worlddata, queen=TRUE) > `summary(queen.nb) > `plot(queen.nb,coords) > > > # the numbers of neighbours of regions in the neighbours list. > > `cards <- card(queen.nb) > > # convert into a listw type based on contiguity row normalized > > `queen.listw <- nb2listw(queen.nb,style="W", zero.policy = TRUE) > `summary(queen.listw,zero.policy=TRUE) > > #marginal effects > `impacts(reg2, listw = listw1,zero.policy=TRUE) > > #marginal effects with p-values > `summary(impacts(reg2, listw = listw1), R= 500, zstats > TRUE,zero.policy=TRUE) > > `attributes(queen.listw) > > #Selection of a subset of countries to fit with mydataset (116 > countries, the previous ones I did for all countries in the shapefile) > > > `setwd("C:/Users/Usuario/Desktop/instituciones/revision energy > economics/spatial panel in stata/countries_shp") > > `worlddata <- readOGR("countries.shp") > > `queen.nb = poly2nb(worlddata, queen=TRUE) > > `coords <-coordinates(worlddata) > > `plot(queen.nb, coords) > > > Following Roger Bivand example: > > `to.be.dropped <- c(1, 2, 5, 6, 8, 9, 10, 13, 14, 18, 20, 21, 23, 26, > 28, 29, 31, 34, 36, 37, 39, 40, 41, 45, 46, 47, 48, 51, 52, 53, 55, 56, 58, > 59, 62, 63, 64, 66, 68, 69, 74, 75, 77, 78, 79, 80, 81, 82, 85, 86, 87, 89, > 90, 94, 95, 97, 98, 99, 100, 102, 103, 104, 105, 107, 109, 110, 118, 123, > 124, 125, 127, 130, 131, 133, 134, 137, 138, 139, 140, 143, 145, 146, 148, > 149, 151, 152, 153, 155, 157, 159, 161, 162, 166, 167, 170, 173, 175, 176, > 177, 178, 181, 182, 183, 184, 186, 190, 193, 194, 195, 198, 199, 200, 201, > 202, 203, 204, 205, 206, 207, 208, 211, 212, 213, 215, 216, 218, 219, 221, > 222, 224, 226, 227, 228, 229, 232, 233, 237, 239, 240, 242, 246, 247, 248, > 255, 258, 259, 260, 261, 265) > sub.queen.nb <- subset(queen.nb, > !(1:length(queen.nb) %in% to.be.dropped)) > which(!(attr(queen.nb, "region.id") %in% > attr(sub.queen.nb, "region.id"))) > > `sub.queen.listw <- nb2listw(sub.queen.nb,style="W", zero.policy > TRUE) > > so now, I have my 116 countries in the queen matrix. > > # My data and variables > `y <- cbind(lei) > `x <- cbind(lgdppcnew, lgdppcnew2, industryg, importsg, fdig) > `xy <- cbind(mydata$x, mydata$y) > `listw1 <- sub.queen.listw > `coords <- coords > > #Define formula > > `reg.ols <- y ~ x > > > # Autocorrelation test > `lm.morantest(ols.eq1,listw1,zero.policy=TRUE) > `plot.moran(ols.eq1,listw1) > > (here I get the error where it is said the objects have different > dimensions) > > `reg2 <- lagsarlm(ols.eq1, data = mydata, listw1, zero.policy=TRUE) > `summary(reg2,correlation=TRUE, Nagelkerke=TRUE, Hausman=TRUE, > zero.policy=TRUE) > > (here I get the error where it is said the objects have different > dimensions) > > #marginal effects > > `impacts(reg2, listw = listw1,zero.policy=TRUE) > > #marginal effects with p-values > > `summary(impacts(reg2, listw = listw1), R= 500, zstats > TRUE,zero.policy=TRUE) > > I don't know if it is for my panel data, the missing values or > something else. > I tried also in Stata having similar issues. > > thanks in advance and my apologies if it is very naive question, but I > am stuck. > > all the best > Maria Jesus > > [[alternative HTML version deleted]] > > ______________________________________________ > R-help at r-project.org mailing list -- To UNSUBSCRIBE and more, see > https://stat.ethz.ch/mailman/listinfo/r-help > PLEASE do read the posting guide > http://www.R-project.org/posting-guide.html > and provide commented, minimal, self-contained, reproducible code. >[[alternative HTML version deleted]]