From noreply at r-forge.r-project.org  Wed Jun 18 17:35:50 2025
From: noreply at r-forge.r-project.org (noreply at r-forge.r-project.org)
Date: Wed, 18 Jun 2025 17:35:50 +0200 (CEST)
Subject: [Gmm-commits] r242 - in pkg/momentfit: . R man
Message-ID: <20250618153550.1B2D01842AD@r-forge.r-project.org>

Author: chaussep
Date: 2025-06-18 17:35:49 +0200 (Wed, 18 Jun 2025)
New Revision: 242

Modified:
   pkg/momentfit/DESCRIPTION
   pkg/momentfit/R/momentData.R
   pkg/momentfit/R/momentModel.R
   pkg/momentfit/R/sysMomentModel-methods.R
   pkg/momentfit/man/gel4.Rd
   pkg/momentfit/man/gmm4.Rd
   pkg/momentfit/man/momentModel.Rd
   pkg/momentfit/man/sysMomentModel.Rd
Log:
fixed a bug for systems of equations with HAC vcov

Modified: pkg/momentfit/DESCRIPTION
===================================================================
--- pkg/momentfit/DESCRIPTION	2024-07-04 19:19:48 UTC (rev 241)
+++ pkg/momentfit/DESCRIPTION	2025-06-18 15:35:49 UTC (rev 242)
@@ -1,8 +1,10 @@
 Package: momentfit
 Version: 1.0
-Date: 2024-05-27
+Date: 2025-06-17
 Title: Methods of Moments
-Author: Pierre Chausse <pchausse at uwaterloo.ca>
+Authors at R: person(given="Pierre", family="Chausse", role=c("aut","cre"),
+	     	    email="pchausse at uwaterloo.ca")
+Author: Pierre Chausse [aut, cre]
 Maintainer: Pierre Chausse <pchausse at uwaterloo.ca>
 Description: Several classes for moment-based models are defined. The classes are defined for moment conditions derived from a single equation or a system of equations. The conditions can also be expressed as functions or formulas. Several methods are also offered to facilitate the development of different estimation techniques. The methods that are currently provided are the Generalized method of moments (Hansen 1982; <doi:10.2307/1912775>), for single equations and systems of equation, and  the Generalized Empirical Likelihood (Smith 1997; <doi:10.1111/j.0013-0133.1997.174.x>, Kitamura 1997; <doi:10.1214/aos/1069362388>, Newey and Smith 2004; <doi:10.1111/j.1468-0262.2004.00482.x>, and Anatolyev 2005 <doi:10.1111/j.1468-0262.2005.00601.x>).Some work is being done to add tools to deal with weak and/or many instruments. This include K-Class estimators (LIML and Fuller), Anderson and Rubin statistics test, etc. 
 Depends: R (>= 3.5), sandwich

Modified: pkg/momentfit/R/momentData.R
===================================================================
--- pkg/momentfit/R/momentData.R	2024-07-04 19:19:48 UTC (rev 241)
+++ pkg/momentfit/R/momentData.R	2025-06-18 15:35:49 UTC (rev 242)
@@ -49,8 +49,9 @@
     {
         modelF <- model.frame(formula, data, na.action="na.pass",
                               drop.unused.levels=TRUE)
+        
         if (is.matrix(modelF[[1]]))
-            return(.multiToSys(formula, h, data))
+            return(.multiToSys(formula, h, data, survOptions, vcovOptions, na.action))
         parNames <- colnames(model.matrix(attr(modelF, "terms"), modelF))
         k <- length(parNames)
         if (any(class(h) == "formula"))

Modified: pkg/momentfit/R/momentModel.R
===================================================================
--- pkg/momentfit/R/momentModel.R	2024-07-04 19:19:48 UTC (rev 241)
+++ pkg/momentfit/R/momentModel.R	2025-06-18 15:35:49 UTC (rev 242)
@@ -130,7 +130,7 @@
                     {
                         model <- .lModelData(g,x,data, survOptions, vcovOptions, na.action)
                         if (!is.null(model$eqnNames))
-                        {                            
+                        {
                             gmodel <- new("slinearModel", data = model$data,
                                           instT=model$instT, 
                                           modelT = model$modelT, vcov = vcov,

Modified: pkg/momentfit/R/sysMomentModel-methods.R
===================================================================
--- pkg/momentfit/R/sysMomentModel-methods.R	2024-07-04 19:19:48 UTC (rev 241)
+++ pkg/momentfit/R/sysMomentModel-methods.R	2025-06-18 15:35:49 UTC (rev 242)
@@ -533,6 +533,7 @@
               sameMom <- object at sameMom
               n <- object at n
               neqn <- length(spec$eqnNames)
+              wSpec <- list()
               if (is.matrix(w))
               {
                   if (!all(dim(w) == sum(spec$q)))
@@ -573,10 +574,13 @@
                   w <- qr(do.call(cbind, gt)/sqrt(n))
                   Sigma <- NULL
               } else {
-                  stop("Only identity, iid and MDS if allowed for now")
+                  w <- vcovHAC(object, theta)
+                  wSpec <- attr(w,"Spec")
+                  type <- "vcov"
+                  Sigma <- NULL
               }
               return(new("sysMomentWeights", type=type,momNames=object at momNames,
-                         wSpec=list(), w=w, Sigma=Sigma, sameMom=sameMom,
+                         wSpec=wSpec, w=w, Sigma=Sigma, sameMom=sameMom,
                          eqnNames=object at eqnNames))
           })
 

Modified: pkg/momentfit/man/gel4.Rd
===================================================================
--- pkg/momentfit/man/gel4.Rd	2024-07-04 19:19:48 UTC (rev 241)
+++ pkg/momentfit/man/gel4.Rd	2025-06-18 15:35:49 UTC (rev 242)
@@ -52,7 +52,7 @@
   respect to \eqn{\theta}.}
 
 \item{vcovOptions}{A list of options for the covariance matrix of the
-  moment conditions. See \code{\link{vcovHAC}} for the default values.}
+  moment conditions. See \code{\link[sandwich]{vcovHAC}} for the default values.}
 
 \item{centeredVcov}{Should the moment function be centered when
   computing its covariance matrix. Doing so may improve inference.}

Modified: pkg/momentfit/man/gmm4.Rd
===================================================================
--- pkg/momentfit/man/gmm4.Rd	2024-07-04 19:19:48 UTC (rev 241)
+++ pkg/momentfit/man/gmm4.Rd	2025-06-18 15:35:49 UTC (rev 242)
@@ -72,7 +72,7 @@
   ready to make that assumption. The option "TrueFixed" is used only
   when the matrix of weights is provided and it is the optimal one. For
   type \code{CL}, clustered covariance matrix is computed. The options
-  are then included in \code{vcovOptions} (see \code{\link{meatCL}}).}
+  are then included in \code{vcovOptions} (see \code{\link[sandwich]{meatCL}}).}
 
   \item{initW}{How should be compute the initial coefficient vector in
     the first. It only makes a difference for linear models for which
@@ -99,7 +99,7 @@
     coefficients. See \code{\link{restModel}} for more details.}
 
   \item{vcovOptions}{A list of options for the covariance matrix of the
-    moment conditions. See \code{\link{vcovHAC}} for the default values.}
+    moment conditions. See \code{\link[sandwich]{vcovHAC}} for the default values.}
 
   \item{survOptions}{If needed, a list with the type of survey weights and
     the weights as a numeric vector, data.frame or formula. The type is either

Modified: pkg/momentfit/man/momentModel.Rd
===================================================================
--- pkg/momentfit/man/momentModel.Rd	2024-07-04 19:19:48 UTC (rev 241)
+++ pkg/momentfit/man/momentModel.Rd	2025-06-18 15:35:49 UTC (rev 242)
@@ -52,10 +52,10 @@
   \code{MDS} only when \code{g} is a formula. In that case, residuals
   are assumed homoscedastic as well as serially uncorrelated. For type
   \code{CL}, clustered covariance matrix is computed. The options are
-  then included in \code{vcovOptions} (see \code{\link{meatCL}}). }
+  then included in \code{vcovOptions} (see \code{\link[sandwich]{meatCL}}). }
 
 \item{vcovOptions}{A list of options for the covariance matrix of the
-  moment conditions. See \code{\link{vcovHAC}} for the default values.}
+  moment conditions. See \code{\link[sandwich]{vcovHAC}} for the default values.}
 
 \item{centeredVcov}{Should the moment function be centered when
   computing its covariance matrix. Doing so may improve inference.}

Modified: pkg/momentfit/man/sysMomentModel.Rd
===================================================================
--- pkg/momentfit/man/sysMomentModel.Rd	2024-07-04 19:19:48 UTC (rev 241)
+++ pkg/momentfit/man/sysMomentModel.Rd	2025-06-18 15:35:49 UTC (rev 242)
@@ -39,10 +39,10 @@
   \code{MDS} only when \code{g} is a formula. In that case, residuals
   are assumed homoscedastic as well as serially uncorrelated. For type
   \code{CL}, clustered covariance matrix is computed. The options are
-  then included in \code{vcovOptions} (see \code{\link{meatCL}}).}
+  then included in \code{vcovOptions} (see \code{\link[sandwich]{meatCL}}).}
 
 \item{vcovOptions}{A list of options for the covariance matrix of the
-  moment conditions. See \code{\link{vcovHAC}} for the default values.}
+  moment conditions. See \code{\link[sandwich]{vcovHAC}} for the default values.}
 
 \item{centeredVcov}{Should the moment function be centered when
   computing its covariance matrix. Doing so may improve inference.}


From noreply at r-forge.r-project.org  Wed Jun 18 20:37:27 2025
From: noreply at r-forge.r-project.org (noreply at r-forge.r-project.org)
Date: Wed, 18 Jun 2025 20:37:27 +0200 (CEST)
Subject: [Gmm-commits] r243 - pkg/momentfit/R
Message-ID: <20250618183727.861F9184BDF@r-forge.r-project.org>

Author: chaussep
Date: 2025-06-18 20:37:27 +0200 (Wed, 18 Jun 2025)
New Revision: 243

Modified:
   pkg/momentfit/R/sysMomentModel-methods.R
Log:
fixed a bug in systems of equations using HAC matrix

Modified: pkg/momentfit/R/sysMomentModel-methods.R
===================================================================
--- pkg/momentfit/R/sysMomentModel-methods.R	2025-06-18 15:35:49 UTC (rev 242)
+++ pkg/momentfit/R/sysMomentModel-methods.R	2025-06-18 18:37:27 UTC (rev 243)
@@ -1025,7 +1025,7 @@
                   else
                       wObj0 <- NULL
                   wObj <- evalWeights(model, theta0, "optimal", wObj0)
-                  if (model at vcov=="HAC" && is.character(model at bw))
+                  if (model at vcov=="HAC" && is.character(model at vcovOptions$bw))
                       model at vcovOptions$bw <- wObj at wSpec$bw
                   res <- solveGmm(model, wObj, theta0, ...)
                   theta1 <- res$theta