start 142

rsbivand · rsbivand · commit 4b4267b6d819 · 2025-09-05T10:22:54.000+02:00
diff --git a/DESCRIPTION b/DESCRIPTION
@@ -1,6 +1,6 @@
 Package: spatialreg
-Version: 1.4-1
-Date: 2025-08-29
+Version: 1.4-2
+Date: 2025-09-05
 Title: Spatial Regression Analysis
 Encoding: UTF-8
 Authors@R: c(person("Roger", "Bivand", role = c("cre", "aut"), email = "Roger.Bivand@nhh.no", comment=c(ORCID="0000-0003-2392-6140")),
@@ -27,7 +27,7 @@ Authors@R: c(person("Roger", "Bivand", role = c("cre", "aut"), email = "Roger.Bi
              person(given = "JuanTomas", family = "Sayago", role = c("ctb"), email = "juantomas.sayago@gmail.com"),
              person("Michael", "Tiefelsdorf", role = "ctb"))
 Depends: R (>= 3.3), spData (>= 2.3.1), Matrix, sf
-Imports: spdep (>= 1.4-1), coda, methods, MASS, boot, splines, LearnBayes,
+Imports: spdep (>= 1.4.1), coda, methods, MASS, boot, splines, LearnBayes,
         nlme, multcomp
 Suggests: parallel, RSpectra, tmap, foreign, spam, knitr, lmtest, expm,
         sandwich, rmarkdown, igraph, tinytest, codingMatrices
diff --git a/NEWS.md b/NEWS.md
@@ -1,4 +1,6 @@
-# Version 1.4-1 (development)
+# Version 1.4-2 (development)
+
+# Version 1.4-1 (2025-09-04)
 
 * import `spdep` >= 1.4.1
 
diff --git a/R/sarlm_tests.R b/R/sarlm_tests.R
@@ -22,7 +22,7 @@ LR.Sarlm <- function(x, y)
 	else LLx <- x
 	if (!inherits(y, "logLik")) LLy <- logLik(y)
 	else LLy <- y
-	statistic <- 2*(LLx - LLy)
+	statistic <- c(2*(LLx - LLy))
 	attr(statistic, "names") <- "Likelihood ratio"
 	parameter <- abs(attr(LLx, "df") - attr(LLy, "df"))
 	if (parameter < 1) 
diff --git a/man/ML_models.Rd b/man/ML_models.Rd
@@ -118,6 +118,8 @@ Refer to the help page of \code{\link{predict.Sarlm}} for discussions and
 references.
 
 Because numerical optimisation is used to find the values of lambda and rho in \code{sacsarlm}, care needs to be shown. It has been found that the surface of the 2D likelihood function often forms a \dQuote{banana trench} from (low rho, high lambda) through (high rho, high lambda) to (high rho, low lambda) values. In addition, sometimes the banana has optima towards both ends, one local, the other global, and conseqently the choice of the starting point for the final optimization becomes crucial. The default approach is not to use just (0, 0) as a starting point, nor the (rho, lambda) values from \code{gstsls}, which lie in a central part of the \dQuote{trench}, but either four values at (low rho, high lambda), (0, 0), (high rho, high lambda), and (high rho, low lambda), and to use the best of these start points for the final optimization. Optionally, nine points can be used spanning the whole (lower, upper) space.
+
+From version 1.4.1, functions for models including spatially lagged independent variables warn on fitting if any of the right-hand side variables are factors. This is because the interpretation of coefficients that are not slopes is unclear when the variable is not interpretable on an unbounded line, such as factors. Factor variable names are shown with the suffix \dQuote{(F)}, others \dQuote{dy/dx} in output from impact methods. A discussion can be found at \url{https://github.com/rsbivand/eqc25_talk}.
 }
 
 \section{Control arguments}{
diff --git a/man/SET_MCMC.Rd b/man/SET_MCMC.Rd
@@ -100,6 +100,9 @@ neighbours, if FALSE (default) assign NA}
     }}
 }}
 
+\details{From version 1.4.1, functions for models including spatially lagged independent variables warn on fitting if any of the right-hand side variables are factors. This is because the interpretation of coefficients that are not slopes is unclear when the variable is not interpretable on an unbounded line, such as factors. Factor variable names are shown with the suffix \dQuote{(F)}, others \dQuote{dy/dx} in output from impact methods. A discussion can be found at \url{https://github.com/rsbivand/eqc25_talk}.
+}
+
 \references{LeSage J and RK Pace (2009) Introduction to Spatial Econometrics. CRC Press, Boca Raton.
 }
 
diff --git a/man/SLX.Rd b/man/SLX.Rd
@@ -54,7 +54,11 @@ for the data on which the model was fitted. Should have row names corresponding
 }
 
 \value{
-The \code{lmSLX} function returns an \dQuote{lm} object with a \dQuote{mixedImps} list of three impact matrixes (impacts and standard errors) for direct, indirect and total impacts; total impacts calculated using a simplified local copy of the estimable function from the gmodels package.
+The \code{lmSLX} function returns an \dQuote{lm} object with a \dQuote{mixedImps} list of three impact matrixes (impacts and standard errors) for direct, indirect and total impacts; total impacts and their standard errors calculated using \code{multcomp::glht}.
+}
+
+
+\details{From version 1.4.1, functions for models including spatially lagged independent variables warn on fitting if any of the right-hand side variables are factors. This is because the interpretation of coefficients that are not slopes is unclear when the variable is not interpretable on an unbounded line, such as factors. Factor variable names are shown with the suffix \dQuote{(F)}, others \dQuote{dy/dx} in output from impact methods. A discussion can be found at \url{https://github.com/rsbivand/eqc25_talk}.
 }
 
 \seealso{\code{\link{lm}}
diff --git a/man/impacts.Rd b/man/impacts.Rd
@@ -66,6 +66,8 @@ If a model is fitted with the \dQuote{Durbin=} set to a formula subsetting the e
 When \code{Q} and \code{tr} are given, addition impact component results are provided for each step in the traces of powers of the weights matrix up to and including the \code{Q}'th power. This increases computing time because the output object is substantially increased in size in proportion to the size of \code{Q}.
 
 The method for \code{gmsar} objects is only for those of \code{type} \code{SARAR} output by \code{gstsls}, and assume that the spatial error coefficient is fixed, and thus omitted from the coefficients and covariance matrix used for simulation.
+
+From version 1.4.1, functions for models including spatially lagged independent variables warn on fitting if any of the right-hand side variables are factors. This is because the interpretation of coefficients that are not slopes is unclear when the variable is not interpretable on an unbounded line, such as factors. Factor variable names are shown with the suffix \dQuote{(F)}, others \dQuote{dy/dx} in output from impact methods. A discussion can be found at \url{https://github.com/rsbivand/eqc25_talk}.
 }
 
 \value{

Original file line number	Diff line number	Diff line change
`@@ -118,6 +118,8 @@ Refer to the help page of \code{\link{predict.Sarlm}} for discussions and`
`118`	`118`	`references.`
`119`	`119`
`120`	`120`	Because numerical optimisation is used to find the values of lambda and rho in \code{sacsarlm}, care needs to be shown. It has been found that the surface of the 2D likelihood function often forms a \dQuote{banana trench} from (low rho, high lambda) through (high rho, high lambda) to (high rho, low lambda) values. In addition, sometimes the banana has optima towards both ends, one local, the other global, and conseqently the choice of the starting point for the final optimization becomes crucial. The default approach is not to use just (0, 0) as a starting point, nor the (rho, lambda) values from \code{gstsls}, which lie in a central part of the \dQuote{trench}, but either four values at (low rho, high lambda), (0, 0), (high rho, high lambda), and (high rho, low lambda), and to use the best of these start points for the final optimization. Optionally, nine points can be used spanning the whole (lower, upper) space.
	`121`	`+`
	`122`	+From version 1.4.1, functions for models including spatially lagged independent variables warn on fitting if any of the right-hand side variables are factors. This is because the interpretation of coefficients that are not slopes is unclear when the variable is not interpretable on an unbounded line, such as factors. Factor variable names are shown with the suffix \dQuote{(F)}, others \dQuote{dy/dx} in output from impact methods. A discussion can be found at \url{https://github.com/rsbivand/eqc25_talk}.
`121`	`123`	`}`
`122`	`124`
`123`	`125`	`\section{Control arguments}{`
Original file line number	Diff line number	Diff line change
`@@ -100,6 +100,9 @@ neighbours, if FALSE (default) assign NA}`
`100`	`100`	`}}`
`101`	`101`	`}}`
`102`	`102`
	`103`	+\details{From version 1.4.1, functions for models including spatially lagged independent variables warn on fitting if any of the right-hand side variables are factors. This is because the interpretation of coefficients that are not slopes is unclear when the variable is not interpretable on an unbounded line, such as factors. Factor variable names are shown with the suffix \dQuote{(F)}, others \dQuote{dy/dx} in output from impact methods. A discussion can be found at \url{https://github.com/rsbivand/eqc25_talk}.
	`104`	`+}`
	`105`	`+`
`103`	`106`	`\references{LeSage J and RK Pace (2009) Introduction to Spatial Econometrics. CRC Press, Boca Raton.`
`104`	`107`	`}`
`105`	`108`
Original file line number	Diff line number	Diff line change
`@@ -54,7 +54,11 @@ for the data on which the model was fitted. Should have row names corresponding`
`54`	`54`	`}`
`55`	`55`
`56`	`56`	`\value{`
`57`		`-The \code{lmSLX} function returns an \dQuote{lm} object with a \dQuote{mixedImps} list of three impact matrixes (impacts and standard errors) for direct, indirect and total impacts; total impacts calculated using a simplified local copy of the estimable function from the gmodels package.`
	`57`	`+The \code{lmSLX} function returns an \dQuote{lm} object with a \dQuote{mixedImps} list of three impact matrixes (impacts and standard errors) for direct, indirect and total impacts; total impacts and their standard errors calculated using \code{multcomp::glht}.`
	`58`	`+}`
	`59`	`+`
	`60`	`+`
	`61`	+\details{From version 1.4.1, functions for models including spatially lagged independent variables warn on fitting if any of the right-hand side variables are factors. This is because the interpretation of coefficients that are not slopes is unclear when the variable is not interpretable on an unbounded line, such as factors. Factor variable names are shown with the suffix \dQuote{(F)}, others \dQuote{dy/dx} in output from impact methods. A discussion can be found at \url{https://github.com/rsbivand/eqc25_talk}.
`58`	`62`	`}`
`59`	`63`
`60`	`64`	`\seealso{\code{\link{lm}}`
Original file line number	Diff line number	Diff line change
`@@ -66,6 +66,8 @@ If a model is fitted with the \dQuote{Durbin=} set to a formula subsetting the e`
`66`	`66`	`When \code{Q} and \code{tr} are given, addition impact component results are provided for each step in the traces of powers of the weights matrix up to and including the \code{Q}'th power. This increases computing time because the output object is substantially increased in size in proportion to the size of \code{Q}.`
`67`	`67`
`68`	`68`	`The method for \code{gmsar} objects is only for those of \code{type} \code{SARAR} output by \code{gstsls}, and assume that the spatial error coefficient is fixed, and thus omitted from the coefficients and covariance matrix used for simulation.`
	`69`	`+`
	`70`	+From version 1.4.1, functions for models including spatially lagged independent variables warn on fitting if any of the right-hand side variables are factors. This is because the interpretation of coefficients that are not slopes is unclear when the variable is not interpretable on an unbounded line, such as factors. Factor variable names are shown with the suffix \dQuote{(F)}, others \dQuote{dy/dx} in output from impact methods. A discussion can be found at \url{https://github.com/rsbivand/eqc25_talk}.
`69`	`71`	`}`
`70`	`72`
`71`	`73`	`\value{`