Merge branch 'dev' into 'master'

Fixes to tests and package checks

Closes #53, #52, and #54

See merge request !33

.gitlab-ci.yml

 # Using the rocker/tidyverse image because it has devtools and testthat.
 
-# Use "document = FALSE" to catch potential mismatches between the roxygen
+# Using "document = FALSE" catches potential mismatches between the roxygen
 # comments and the .Rd files. This could happen when checking locally
 # with "document = TRUE" (the default), which updates the .Rd files, but then
 # forgetting to commit and push those updated .Rd files.
 
+# Setting the environment variable "_R_CHECK_SYSTEM_CLOCK_" to 0
+# to disable package time check, see issue #54
+
+# Run the R commands under a single "R -e" so everything runs in
+# the same R session.
+# Note for the future: If there are going to be more R commands just put them
+# in a separate R script and run that from here.
+
 .check_template: &check_job_template
   only:
     - master
     - merge_requests
     - tags
   script:
-    - R -e 'sessionInfo()'
-    - R -e 'devtools::install_deps(quiet = TRUE)'
-    - R -e 'devtools::check(error_on = "note", document = FALSE)'
+    - R -e
+        "sessionInfo();
+        Sys.setenv('_R_CHECK_SYSTEM_CLOCK_' = 0);
+        devtools::install_deps(quiet = TRUE);
+        devtools::check(error_on = 'note', document = FALSE)"
 
 # Earliest supported version
 check-3.5.1:

DESCRIPTION

@@ -2,11 +2,11 @@ Package: stressaddition
 Type: Package
 Title: Modelling Tri-Phasic Concentration-Response Relationships
 Version: 3.0.2
-Date: 2020-08-18
+Date: 2020-09-04
 Authors@R: c(person("Sebastian", 
                     "Henz", 
                     role = c("aut", "cre"), 
-                    email = "sebastian.henz@ufz.de", 
+                    email = "bastihz.dev@posteo.de", 
                     comment = c(ORCID = "0000-0001-8299-8852")),
              person("Matthias",
                     "Liess",

NEWS.md

-# stressaddition (development version)
+# stressaddition 3.0.2
 
 * Fixed a bug where the plotting functions printed `NULL` to the console.
+* Changed maintainer e-mail address to ensure long-term maintainability.
+* Internal improvements related to package testing.
 
 
 # stressaddition 3.0.1
+
 * Added missing return values in documentation to comply with CRAN repository policy.
 * Added a link to the paper in the description file.
 

cran-comments.md

+## Change in the maintainer's email address
+I recently learned that my old address (sebastian.henz@ufz.de) is probably 
+going to get deactivated in a few months. To ensure long-term maintainability 
+of the package I changed the address to one that will remain active for the 
+foreseeable future. I have sent a confirmation from the old address to
+CRAN-submissions@R-project.org.
+
 ## Test environments
-* local Windows 10 install, R 4.0.2
-* win-builder, R devel
-* Debian GNU/Linux 9 on GitLab CI, R 3.5.1
-* Debian GNU/Linux 10 on GitLab CI, R 3.6.3
-* Ubuntu 20.04 on GitLab CI, R 4.0.2
-* Ubuntu 20.04 on GitLab CI, R devel (2020-07-21 r78893)
+* local Windows 10 installation, R 4.0.2
+* Ubuntu Linux 16.04 LTS on R-hub, R-release
+* Windows Server 2008 R2 SP1 on R-hub, R-devel
+* Ubuntu 20.04 LTS on GitLab-CI, R 4.0.2
+* Ubuntu 20.04 LTS on GitLab-CI, R devel (2020-08-18 r79041)
+* Debian GNU/Linux 10 on GitLab-CI, R 3.6.3
+* Debian GNU/Linux 9 on GitLab-CI, R 3.5.1
 
 ## R CMD check results
-There were no errors, warnings, or notes. 
+There were no errors or warnings.
+One note came up: "unable to verify current time". I believe this is 
+not caused by my package but instead by how R checks timestamps.
 
 ## Downstream dependencies
 There are no downstream dependencies yet.

man/stressaddition-package.Rd

@@ -37,7 +37,7 @@ Useful links:
 
 }
 \author{
-\strong{Maintainer}: Sebastian Henz \email{sebastian.henz@ufz.de} (\href{https://orcid.org/0000-0001-8299-8852}{ORCID})
+\strong{Maintainer}: Sebastian Henz \email{bastihz.dev@posteo.de} (\href{https://orcid.org/0000-0001-8299-8852}{ORCID})
 
 Authors:
 \itemize{

tests/testthat/test-lc.R

@@ -50,45 +50,41 @@ test_that("all input formats produce identical models", {
 
 
 test_that("lc values have not changed", {
+    # Don't test too accurately because somehow testing on different
+    # operating systems may give slightly different results.
+    # Remember to use rhub for running tests on all major platforms
+    # including macos.
     expect_equal(
-        lc(model, "survival_tox_sys", 50),
-        list(response = 44.95368, concentration = 3.375735),
-        tolerance = 1e-4
+        lapply(lc(model, "survival_tox_sys", 50), round, 2),
+        list(response = 44.95, concentration = 3.38)
     )
     expect_equal(
-        lc(model, "survival_tox_sys", 10),
-        list(response = 80.91662, concentration = 1.098648),
-        tolerance = 1e-4
+        lapply(lc(model, "survival_tox_sys", 10), round, 2),
+        list(response = 80.92, concentration = 1.10)
     )
     expect_equal(
-        lc(model, "survival_tox", 100/3),
-        list(response = 66.66667, concentration = 1.902125),
-        tolerance = 1e-4
+        lapply(lc(model, "survival_tox", 100/3), round, 2),
+        list(response = 66.67, concentration = 1.90)
     )
     expect_equal(
-        lc(model, "survival_tox_LL5", 42),
-        list(response = 52.2, concentration = 2.054426),
-        tolerance = 1e-4
+        lapply(lc(model, "survival_tox_LL5", 42), round, 2),
+        list(response = 52.20, concentration = 2.05)
     )
     expect_equal(
-        lc(model, "survival_tox_env_sys", 50),
-        list(response = 14.67725, concentration = 1.299516),
-        tolerance = 1e-4
+        lapply(lc(model, "survival_tox_env_sys", 50), round, 2),
+        list(response = 14.68, concentration = 1.30)
     )
     expect_equal(
-        lc(model, "survival_tox_env_sys", 10),
-        list(response = 26.41904, concentration = 0.0008571244),
-        tolerance = 1e-4
+        lapply(lc(model, "survival_tox_env_sys", 10), round, 4),
+        list(response = 26.419, concentration = 0.0009)
     )
     expect_equal(
-        lc(model, "survival_tox_env", 67.89),
-        list(response = 24.51453, concentration = 0.7890892),
-        tolerance = 1e-4
+        lapply(lc(model, "survival_tox_env", 67.89), round, 2),
+        list(response = 24.51, concentration = 0.79)
     )
     expect_equal(
-        lc(model, "survival_tox_env_LL5", 3.14159),
-        list(response = 28.73466, concentration = 0.7267524),
-        tolerance = 1e-4
+        lapply(lc(model, "survival_tox_env_LL5", 3.14), round, 2),
+        list(response = 28.74, concentration = 0.73)
     )
 })