From 8de6e0c9df935266eccebbc9ff8ff2eef2e819d1 Mon Sep 17 00:00:00 2001 From: Johannes Ranke Date: Thu, 17 Dec 2020 08:45:07 +0100 Subject: Remove two skipped (on Travis) tests to reduce noise --- tests/testthat/test_mkinfit_errors.R | 16 ---------------- 1 file changed, 16 deletions(-) (limited to 'tests/testthat/test_mkinfit_errors.R') diff --git a/tests/testthat/test_mkinfit_errors.R b/tests/testthat/test_mkinfit_errors.R index 85ee574e..4c8177bd 100644 --- a/tests/testthat/test_mkinfit_errors.R +++ b/tests/testthat/test_mkinfit_errors.R @@ -35,19 +35,3 @@ test_that("mkinfit warns if a specified initial parameter value is not in the mo expect_warning(mkinfit("SFO", FOCUS_2006_A, parms.ini = c(k_xy = 0.1), quiet = TRUE), "not used in the model") }) - -test_that("We get reproducible output if quiet = FALSE", { - # We cannot expect parameter and sum of squares traces to be the same across platforms - skip_on_cran() - skip_on_travis() - expect_known_output(mkinfit("DFOP", FOCUS_2006_C, reweight.method = "tc", trace_parms = TRUE), - file = "DFOP_FOCUS_C_messages.txt") -}) - -test_that("We get warnings in case of overparameterisation", { - skip_on_cran() # On winbuilder the following fit does not give a warning - skip_on_travis() # Neither on travis - expect_warning(f <- mkinfit("FOMC", FOCUS_2006_A, quiet = TRUE), "not converge") - # We do get Hessians and the related output after the switch to using numDeriv::hessian() - #s2 <- expect_warning(summary(mkinfit("DFOP", FOCUS_2006_A, quiet = TRUE)), "singular system") -}) -- cgit v1.2.3