1 context("computeMedoidsIndices")
3 test_that("serialization + getDataInFile retrieve original data / from matrix",
5 data_bin_file = "/tmp/epclust_test_m.bin"
8 #dataset 200 lignes / 30 columns
9 data_ascii = matrix(runif(200*30,-10,10),ncol=30)
10 nbytes = 4 #lead to a precision of 1e-7 / 1e-8
13 #Simulate serialization in one single call
14 binarize(data_ascii, data_bin_file, 500, ",", nbytes, endian)
15 expect_equal(file.info(data_bin_file)$size, length(data_ascii)*nbytes+8)
16 for (indices in list(c(1,3,5), 3:13, c(5,20,50), c(75,130:135), 196:200))
18 data_lines = getDataInFile(indices, data_bin_file, nbytes, endian)
19 expect_equal(data_lines, data_ascii[indices,], tolerance=1e-6)
23 #...in several calls (last call complete, next call NULL)
25 binarize(data_ascii[((i-1)*10+1):(i*10),], data_bin_file, 20, ",", nbytes, endian)
26 expect_equal(file.info(data_bin_file)$size, length(data_ascii)*nbytes+8)
27 for (indices in list(c(1,3,5), 3:13, c(5,20,50), c(75,130:135), 196:200))
29 data_lines = getDataInFile(indices, data_bin_file, nbytes, endian)
30 expect_equal(data_lines, data_ascii[indices,], tolerance=1e-6)
35 TODO: test computeMedoids + filter
36 # #R-equivalent, requiring a matrix (thus potentially breaking "fit-in-memory" hope)
37 # mat_meds = medoids[,]
38 # mi = rep(NA,nb_series)
39 # for (i in 1:nb_series)
40 # mi[i] <- which.min( rowSums( sweep(mat_meds, 2, ref_series[i,], '-')^2 ) )