epclust/tests/testthat/test.clustering.R

   1 context("clustering")
   2
   3 #shorthand: map 1->1, 2->2, 3->3, 4->1, ..., 149->2, 150->3, ... (is base==3)
   4 I = function(i, base)
   5     (i-1) %% base + 1
   6
   7 test_that("computeClusters1&2 behave as expected",
   8 {
   9     require("MASS", quietly=TRUE)
  10     if (!require("clue", quietly=TRUE))
  11         skip("'clue' package not available")
  12
  13     # 3 gaussian clusters, 300 items; and then 7 gaussian clusters, 490 items
  14     n = 300
  15     d = 5
  16     K = 3
  17     for (ndK in list( c(300,5,3), c(490,10,7) ))
  18     {
  19         n = ndK[1] ; d = ndK[2] ; K = ndK[3]
  20         cs = n/K #cluster size
  21         Id = diag(d)
  22         coefs = sapply(1:K, function(i) MASS::mvrnorm(cs, c(rep(0,(i-1)),5,rep(0,d-i)), Id))
  23         indices_medoids1 = computeClusters1(coefs, K, verbose=TRUE)
  24         indices_medoids2 = computeClusters2(dist(coefs), K, verbose=TRUE)
  25         # Get coefs assignments (to medoids)
  26         assignment1 = sapply(seq_len(n), function(i)
  27             which.min( colSums( sweep(coefs[,indices_medoids1],1,coefs[,i],'-')^2 ) ) )
  28         assignment2 = sapply(seq_len(n), function(i)
  29             which.min( colSums( sweep(coefs[,indices_medoids2],1,coefs[,i],'-')^2 ) ) )
  30         for (i in 1:K)
  31         {
  32             expect_equal(sum(assignment1==i), cs, tolerance=5)
  33             expect_equal(sum(assignment2==i), cs, tolerance=5)
  34         }
  35
  36         costs_matrix1 = matrix(nrow=K,ncol=K)
  37         costs_matrix2 = matrix(nrow=K,ncol=K)
  38         for (i in 1:K)
  39         {
  40             for (j in 1:K)
  41             {
  42                 # assign i (in result) to j (order 1,2,3)
  43                 costs_matrix1[i,j] = abs( mean(assignment1[((i-1)*cs+1):(i*cs)]) - j )
  44                 costs_matrix2[i,j] = abs( mean(assignment2[((i-1)*cs+1):(i*cs)]) - j )
  45             }
  46         }
  47         permutation1 = as.integer( clue::solve_LSAP(costs_matrix1) )
  48         permutation2 = as.integer( clue::solve_LSAP(costs_matrix2) )
  49         for (i in 1:K)
  50         {
  51             expect_equal(
  52                 mean(assignment1[((i-1)*cs+1):(i*cs)]), permutation1[i], tolerance=0.05)
  53             expect_equal(
  54                 mean(assignment2[((i-1)*cs+1):(i*cs)]), permutation2[i], tolerance=0.05)
  55         }
  56     }
  57 })
  58
  59 test_that("computeSynchrones behave as expected",
  60 {
  61     n = 300
  62     x = seq(0,9.5,0.1)
  63     L = length(x) #96 1/4h
  64     K = 3
  65     s1 = cos(x)
  66     s2 = sin(x)
  67     s3 = c( s1[1:(L%/%2)] , s2[(L%/%2+1):L] )
  68     #sum((s1-s2)^2) == 96
  69     #sum((s1-s3)^2) == 58
  70     #sum((s2-s3)^2) == 38
  71     s = list(s1, s2, s3)
  72     series = matrix(nrow=L, ncol=n)
  73     for (i in seq_len(n))
  74         series[,i] = s[[I(i,K)]] + rnorm(L,sd=0.01)
  75     getRefSeries = function(indices) {
  76         indices = indices[indices <= n]
  77         if (length(indices)>0) series[,indices] else NULL
  78     }
  79     synchrones = computeSynchrones(bigmemory::as.big.matrix(cbind(s1,s2,s3)), getRefSeries,
  80         n, 100, verbose=TRUE, parll=FALSE)
  81
  82     expect_equal(dim(synchrones), c(L,K))
  83     for (i in 1:K)
  84         expect_equal(synchrones[,i], s[[i]], tolerance=0.01)
  85 })
  86
  87 # NOTE: medoids can be a big.matrix
  88 computeDistortion = function(series, medoids)
  89 {
  90     n = nrow(series) ; L = ncol(series)
  91     distortion = 0.
  92     if (bigmemory::is.big.matrix(medoids))
  93         medoids = medoids[,]
  94     for (i in seq_len(n))
  95         distortion = distortion + min( colSums( sweep(medoids,1,series[,i],'-')^2 ) / L )
  96     distortion / n
  97 }
  98
  99 test_that("clusteringTask1 behave as expected",
 100 {
 101     n = 900
 102     x = seq(0,9.5,0.1)
 103     L = length(x) #96 1/4h
 104     K1 = 60
 105     s = lapply( seq_len(K1), function(i) x^(1+i/30)*cos(x+i) )
 106     series = matrix(nrow=n, ncol=L)
 107     for (i in seq_len(n))
 108         series[,i] = s[[I(i,K1)]] + rnorm(L,sd=0.01)
 109     getSeries = function(indices) {
 110         indices = indices[indices <= n]
 111         if (length(indices)>0) series[,indices] else NULL
 112     }
 113     wf = "haar"
 114     ctype = "absolute"
 115     getContribs = function(indices) curvesToContribs(series[,indices],wf,ctype)
 116     indices1 = clusteringTask1(1:n, getContribs, K1, 75, verbose=TRUE, parll=FALSE)
 117     medoids_K1 = getSeries(indices1)
 118
 119     expect_equal(dim(medoids_K1), c(L,K1))
 120     # Not easy to evaluate result: at least we expect it to be better than random selection of
 121     # medoids within initial series
 122     distorGood = computeDistortion(series, medoids_K1)
 123     for (i in 1:3)
 124         expect_lte( distorGood, computeDistortion(series,series[,sample(1:n, K1)]) )
 125 })
 126
 127 test_that("clusteringTask2 behave as expected",
 128 {
 129     n = 900
 130     x = seq(0,9.5,0.1)
 131     L = length(x) #96 1/4h
 132     K1 = 60
 133     K2 = 3
 134     #for (i in 1:60) {plot(x^(1+i/30)*cos(x+i),type="l",col=i,ylim=c(-50,50)); par(new=TRUE)}
 135     s = lapply( seq_len(K1), function(i) x^(1+i/30)*cos(x+i) )
 136     series = matrix(nrow=n, ncol=L)
 137     for (i in seq_len(n))
 138         series[i,] = s[[I(i,K1)]] + rnorm(L,sd=0.01)
 139     getRefSeries = function(indices) {
 140         indices = indices[indices <= n]
 141         if (length(indices)>0) series[,indices] else NULL
 142     }
 143     # Artificially simulate 60 medoids - perfect situation, all equal to one of the refs
 144     medoids_K1 = bigmemory::as.big.matrix( sapply( 1:K1, function(i) s[[I(i,K1)]] ) )
 145     medoids_K2 = clusteringTask2(medoids_K1, K2, getRefSeries, n, 75, verbose=TRUE, parll=FALSE)
 146
 147     expect_equal(dim(medoids_K2), c(L,K2))
 148     # Not easy to evaluate result: at least we expect it to be better than random selection of
 149     # medoids within 1...K1 (among references)
 150     distorGood = computeDistortion(series, medoids_K2)
 151     for (i in 1:3)
 152         expect_lte( distorGood, computeDistortion(series,medoids_K1[,sample(1:K1, K2)]) )
 153 })
 154
 155 #NOTE: rather redundant test
 156 #test_that("clusteringTask1 + clusteringTask2 behave as expected",
 157 #{
 158 #   n = 900
 159 #   x = seq(0,9.5,0.1)
 160 #   L = length(x) #96 1/4h
 161 #   K1 = 60
 162 #   K2 = 3
 163 #   s = lapply( seq_len(K1), function(i) x^(1+i/30)*cos(x+i) )
 164 #   series = matrix(nrow=n, ncol=L)
 165 #   for (i in seq_len(n))
 166 #       series[i,] = s[[I(i,K1)]] + rnorm(L,sd=0.01)
 167 #   getSeries = function(indices) {
 168 #       indices = indices[indices <= n]
 169 #       if (length(indices)>0) series[indices,] else NULL
 170 #   }
 171 #   wf = "haar"
 172 #   ctype = "absolute"
 173 #   getContribs = function(indices) curvesToContribs(series[indices,],wf,ctype)
 174 #   require("bigmemory", quietly=TRUE)
 175 #   indices1 = clusteringTask1(1:n, getContribs, K1, 75, verbose=TRUE, parll=FALSE)
 176 #   medoids_K1 = bigmemory::as.big.matrix( getSeries(indices1) )
 177 #   medoids_K2 = clusteringTask2(medoids_K1, K2, getSeries, n, 120, verbose=TRUE, parll=FALSE)
 178 #
 179 #   expect_equal(dim(medoids_K1), c(K1,L))
 180 #   expect_equal(dim(medoids_K2), c(K2,L))
 181 #   # Not easy to evaluate result: at least we expect it to be better than random selection of
 182 #   # medoids within 1...K1 (among references)
 183 #   distorGood = computeDistortion(series, medoids_K2)
 184 #   for (i in 1:3)
 185 #       expect_lte( distorGood, computeDistortion(series,medoids_K1[sample(1:K1, K2),]) )
 186 #})