submitted version 0.2-5

DESCRIPTION

@@ -1,40 +1,40 @@
 Package: PlackettLuce
 Type: Package
 Title: Plackett-Luce Models for Rankings
-Version: 0.2-4
-Authors@R: c(person("Heather", "Turner", 
+Version: 0.2-5
+Authors@R: c(person("Heather", "Turner",
         email = "ht@heatherturner.net", role = c("aut", "cre"),
         comment = c(ORCID = "0000-0002-1256-3375")),
         person("Ioannis", "Kosmidis", role = "aut",
-        comment = c(ORCID = "0000-0003-1556-0302")), 
-        person("David", "Firth", role = "aut", 
-        comment = c(ORCID = "0000-0003-0302-2312")), 
+        comment = c(ORCID = "0000-0003-1556-0302")),
+        person("David", "Firth", role = "aut",
+        comment = c(ORCID = "0000-0003-0302-2312")),
         person("Jacob", "van Etten", role = "ctb",
         comment = c(ORCID = "0000-0001-7554-2558")))
 URL: https://hturner.github.io/PlackettLuce/
 BugReports: https://github.com/hturner/PlackettLuce/issues
 Description: Functions to prepare rankings data and fit the Plackett-Luce model
-    jointly attributed to Plackett (1975) <doi:10.2307/2346567> and Luce 
-    (1959, ISBN:0486441369). The standard Plackett-Luce model is generalized 
-    to accommodate ties of any order in the ranking. Partial rankings, in which 
+    jointly attributed to Plackett (1975) <doi:10.2307/2346567> and Luce
+    (1959, ISBN:0486441369). The standard Plackett-Luce model is generalized
+    to accommodate ties of any order in the ranking. Partial rankings, in which
     only a subset of items are ranked in each ranking, are also accommodated in
-    the implementation. Disconnected/weakly connected networks implied by the 
-    rankings may be handled by adding pseudo-rankings with a hypothetical item. 
-    Optionally, a multivariate normal prior may be set on the log-worth 
-    parameters and ranker reliabilities may be incorporated as proposed by 
-    Raman and Joachims (2014) <doi:10.1145/2623330.2623654>. Maximum a 
-    posteriori estimation is used when priors are set. Methods are provided to 
-    estimate standard errors or quasi-standard errors for inference as well as 
-    to fit Plackett-Luce trees. See the package website or vignette for further 
+    the implementation. Disconnected/weakly connected networks implied by the
+    rankings may be handled by adding pseudo-rankings with a hypothetical item.
+    Optionally, a multivariate normal prior may be set on the log-worth
+    parameters and ranker reliabilities may be incorporated as proposed by
+    Raman and Joachims (2014) <doi:10.1145/2623330.2623654>. Maximum a
+    posteriori estimation is used when priors are set. Methods are provided to
+    estimate standard errors or quasi-standard errors for inference as well as
+    to fit Plackett-Luce trees. See the package website or vignette for further
     details.
 License: GPL-3
 Encoding: UTF-8
 LazyData: true
 Depends: R (>= 2.10)
-Imports: Matrix, igraph, methods, partykit, psychotools, 
+Imports: Matrix, igraph, methods, partykit, psychotools,
     psychotree, rARPACK, qvcalc, sandwich, stats
-Suggests: BiocStyle, BradleyTerry2, BradleyTerryScalable, 
-    Matrix.utils, PLMIX, StatRank, covr, hyper2, kableExtra, knitr, 
+Suggests: BiocStyle, BradleyTerry2, BradleyTerryScalable,
+    Matrix.utils, PLMIX, StatRank, covr, hyper2, kableExtra, knitr,
     lbfgs, gnm, pmr, rmarkdown, testthat, tibble
 RoxygenNote: 6.1.1
 VignetteBuilder: knitr

NEWS.md

@@ -1,3 +1,7 @@
+# PlackettLuce 0.2-5
+
+* Higher tolerance in tests of `vcov()` for CRAN Windows test machine.
+
 # PlackettLuce 0.2-4
 
 ## New Features

cran-comments.md

@@ -1,21 +1,16 @@
 ## Comments
 
-This update addresses the errors resulting from the change of method for 
-generating from a discrete uniform distribution.
+This is a re-submission to fix the failed tests on r-devel-windows-ix86+x86_64.
 
 ## Test environments
-
-* Local
- - Ubuntu 18.04.2 LTS, R 3.5.3
- - Ubuntu 18.04, R Under development (unstable) (2019-03-19 r76252)
- - Windows 8, R 3.5.3
 
-* Via R-hub 
- - Mac OS 10.11 El Capitan, R-release (experimental)
+* Via Win-builder
+ - Windows Server 2008 (64-bit), R-devel
 
 ### Check results
 
-I get one warning and one note on R-hub macOS.
+The test environment replicated the error with version 0.2-4.
+
+This submission (0.2-5), passes with no error/warnings.
 
-* The warning `pandoc: Could not fetch https://www.r-pkg.org/badges/version/PlackettLuce` is due to missing https support in an older version of pandoc.
-* The note `Found the following (possibly) invalid URL/DOIs` is a false alarm, I have checked the URLs work and the DOIs resolve (all go to jstor.org).
+There remains a note regarding possibly invalid URLs/DOIs which I believe is a false alarm as I have checked these URLs/DOIs. It may be due to a captcha challenge. They are all links to jstor.org.

tests/testthat/test-vcov.R

@@ -94,7 +94,7 @@ test_that("vcov.PlackettLuce approximated by vcov_hessian [gamma prior]", {
     # should be equal to low tolerance if base on observed Fisher Info though
     expect_equal(vcov(gamma_prior, type = "observed"),
                  vcov_hessian(gamma_prior),
-                 check.attributes = FALSE, tol = 1e-7)
+                 check.attributes = FALSE, tol = 1e-6)
     # more data - N.B. this assumes each ranker gives exactly same ranking,
     # not likely in practice just checking statistical property here
     # (vcov_hessian slow with large number of adherence par, not practical to
@@ -107,7 +107,7 @@ test_that("vcov.PlackettLuce approximated by vcov_hessian [gamma prior]", {
                  check.attributes = FALSE, tol = 1e-3)
     expect_equal(vcov(gamma_prior, type = "observed"),
                  vcov_hessian(gamma_prior),
-                 check.attributes = FALSE, tol = 1e-7)
+                 check.attributes = FALSE, tol = 1e-6)
 })
 
 test_that("vcov.PlackettLuce works, grouped rankings [normal + gamma prior]", {
@@ -128,7 +128,7 @@ test_that("vcov.PlackettLuce works, grouped rankings [normal + gamma prior]", {
     # but that based on observed info equals numerical hessian to medium tol
     expect_equal(vcov(both_priors, type = "observed"),
                  vcov_hessian(both_priors),
-                 check.attributes = FALSE, tol = 1e-7)
+                 check.attributes = FALSE, tol = 1e-6)
     # gamma prior only (different method for Info inversion)
     gamma_prior <- PlackettLuce(rankings = G, npseudo = 0, method = "BFGS",
                                 gamma = list(shape = 100, rate = 100))