# Both tests check whether most of the clustered feature have been found (50% of the 500 features) test_that("vsclust_algorithm", { data("artificial_clusters") dat <- averageCond(artificial_clusters, 5, 10) dat <- scale(dat) clust_out <- vsclust_algorithm(dat, centers = 6, m = 1.55) expect_equal( sum(apply(clust_out$membership, 1, max) > 0.5), 200, tolerance = 10) }) test_that("clust_comp", { data("artificial_clusters") dat <- averageCond(artificial_clusters, 5, 10) clust_out <- ClustComp(dat, NClust = 6, Sds = 1) sum(apply(clust_out$Bestcl$membership, 1, max) > 0.5) expect_equal( sum(apply(clust_out$Bestcl$membership, 1, max) > 0.5), 214, tolerance = 10) }) test_that("sign_analysis_paired", { set.seed(0) data <- matrix(rnorm(1:1500), nrow=100) # make artificial regulations data[1:20,c(2,5,8,11,14)] <- data[1:20, c(1,4,7,10,13)] + 3 data[21:40,c(2,5,8,11,14)] <- data[21:40, c(1,4,7,10,13)] - 3 # Run statistical testing stat_out <- SignAnalysisPaired(data, 3, 5) # Histogram of qvalues comparing the second to the first condition expect_equal(sum(stat_out$qvalues[,1] < 0.01), 42) }) test_that("sign_analysis_unpaired", { set.seed(1) data <- matrix(rnorm(1:1500), nrow=100) # make artificial regulations data[1:20,c(2,5,8,11,14)] <- data[1:20,c(2,5,8,11,14)] + 3 data[21:40,c(2,5,8,11,14)] <- data[21:40,c(2,5,8,11,14)] - 3 # Run statistical testing stat_out <- SignAnalysis(data, 3, 5) # Histogram of qvalues comparing the second to the first condition expect_equal(sum(stat_out$qvalues[,1] < 0.01), 40) }) test_that("prepare_for_vsclust", { set.seed(1) data <- matrix(rnorm(1:1500), nrow=100) # make artificial regulations data[1:20,c(2,5,8,11,14)] <- data[1:20,c(2,5,8,11,14)] + 3 data[21:40,c(2,5,8,11,14)] <- data[21:40,c(2,5,8,11,14)] - 3 # Run statistical testing prepared <- PrepareForVSClust(data, 5,3, isStat=TRUE) expect_equal(sum(prepared$statFileOut[,"qvalue BvsA"] < 0.01), 40) }) test_that("estimate_clust_num", { set.seed(0) data("artificial_clusters") dat <- averageCond(artificial_clusters, 5, 10) dat <- scale(dat) dat <- cbind(dat, 1) ClustInd <- estimClustNum(dat, 10) expect_equal(as.numeric(optimalClustNum(ClustInd, index="MinCentroidDist", method="VSClust")), 6) })