X-Git-Url: https://git.auder.net/?p=morpheus.git;a=blobdiff_plain;f=pkg%2Ftests%2Ftestthat%2Ftest-computeMu.R;h=c1fd4f910e5c63fae03499eec47e70244694cd0a;hp=fccef55fbfa3f49841a69ae575642b050d2eedd4;hb=ab35f6102896a49e86e853262c0650faa2931638;hpb=6dd5c2acccd10635449230faa824b7e8906911bf

diff --git a/pkg/tests/testthat/test-computeMu.R b/pkg/tests/testthat/test-computeMu.R
index fccef55..c1fd4f9 100644
--- a/pkg/tests/testthat/test-computeMu.R
+++ b/pkg/tests/testthat/test-computeMu.R
@@ -1,23 +1,21 @@
-context("computeMu")
-
 test_that("on input of sufficient size, β/||β|| is estimated accurately enough",
 {
-  n = 100000
-  d = 2
-  K = 2
-  p = 1/2
+  n <- 100000
+  d <- 2
+  K <- 2
+  p <- 1/2
 
-  βs_ref = array( c(1,0,0,1 , 1,-2,3,1), dim=c(d,K,2) )
+  βs_ref <- array( c(1,0,0,1 , 1,-2,3,1), dim=c(d,K,2) )
   for (i in 1:(dim(βs_ref)[3]))
   {
-    μ_ref = normalize(βs_ref[,,i])
-    for (model in c("logit","probit"))
+    μ_ref <- normalize(βs_ref[,,i])
+    for (link in c("logit","probit"))
     {
-      cat("\n\n",model," :\n",sep="")
+      cat("\n\n",link," :\n",sep="")
 
-      io = generateSampleIO(n, p, βs_ref[,,i], rep(0,K), model)
-      μ = computeMu(io$X, io$Y, list(K=K))
-      μ_aligned = alignMatrices(list(μ), ref=μ_ref, ls_mode="exact")[[1]]
+      io <- generateSampleIO(n, p, βs_ref[,,i], rep(0,K), link)
+      μ <- computeMu(io$X, io$Y, list(K=K))
+      μ_aligned <- alignMatrices(list(μ), ref=μ_ref, ls_mode="exact")[[1]]
 
       #Some traces: 0 is not well estimated, but others are OK
       cat("Reference normalized matrix:\n")
@@ -25,11 +23,11 @@ test_that("on input of sufficient size, β/||β|| is estimated accurately enough
       cat("Estimated normalized matrix:\n")
       print(μ_aligned)
       cat("Difference norm (Matrix norm ||.||_1, max. abs. sum on a column)\n")
-      diff_norm = norm(μ_ref - μ_aligned)
+      diff_norm <- norm(μ_ref - μ_aligned)
       cat(diff_norm,"\n")
 
-      #NOTE: 0.5 is loose threshold, but values around 0.3 are expected...
-      expect_that( diff_norm, is_less_than(0.5) )
+      # NOTE: 0.5 is loose threshold, but values around 0.3 are expected...
+      expect_lt( diff_norm, 0.5 )
     }
   }
 })