de_serialize works. Variables names improved. Code beautified. TODO: clustering tests

[epclust.git] / epclust / tests / testthat / test.de_serialize.R

diff --git a/epclust/tests/testthat/test.de_serialize.R b/epclust/tests/testthat/test.de_serialize.R
index 41c57e4..27e6d59 100644 (file)
--- a/epclust/tests/testthat/test.de_serialize.R
+++ b/epclust/tests/testthat/test.de_serialize.R
@@ -1,52 +1,65 @@
+context("de_serialize")
 
-#data: matrix of double or connection
-serialize = function(data, file, type, nb_per_chunk)
+data_bin_file <<- "/tmp/epclust_test.bin"
+unlink(data_bin_file)
+
+test_that("serialization + getDataInFile retrieve original data / from matrix",
 {
-       bin_data = file(file, "ab")
-       #write data length on first call
-       nbytes = ifelse(type=="double",8,4)
-       first_write = FALSE
-       if (file.info(file)$size == 0)
+       #dataset 200 lignes / 30 columns
+       data_ascii = matrix(runif(200*30,-10,10),ncol=30)
+       nbytes = 4 #lead to a precision of 1e-7 / 1e-8
+       endian = "little"
+
+       #Simulate serialization in one single call
+       serialize(data_ascii, data_bin_file, 500, ",", nbytes, endian)
+       expect_equal(file.info(data_bin_file)$size, length(data_ascii)*nbytes+8)
+       for (indices in list(c(1,3,5), 3:13, c(5,20,50), c(75,130:135), 196:200))
        {
-               #number of items always on 8 bytes
-               writeBin(0L, bin_data, size=8) #,endian="little")
-               first_write = TRUE
+               data_lines = getDataInFile(indices, data_bin_file, nbytes, endian)
+               expect_equal(data_lines, data_ascii[indices,], tolerance=1e-6)
        }
-       if (is.matrix(data))
+       unlink(data_bin_file)
+
+       #...in several calls (last call complete, next call NULL)
+       for (i in 1:20)
+               serialize(data_ascii[((i-1)*10+1):(i*10),], data_bin_file, 20, ",", nbytes, endian)
+       expect_equal(file.info(data_bin_file)$size, length(data_ascii)*nbytes+8)
+       for (indices in list(c(1,3,5), 3:13, c(5,20,50), c(75,130:135), 196:200))
        {
-               writeBin(t(data), bin_data, size=nbytes)
-               data_length = ncol(data)
+               data_lines = getDataInFile(indices, data_bin_file, nbytes, endian)
+               expect_equal(data_lines, data_ascii[indices,], tolerance=1e-6)
        }
-       else #if (is(data, "connection"))
+       unlink(data_bin_file)
+})
+
+test_that("serialization + getDataInFile retrieve original data / from connection",
+{
+       #dataset 300 lignes / 50 columns
+       data_csv = system.file("testdata","de_serialize.csv",package="epclust")
+       nbytes = 8
+       endian = "big"
+       data_ascii = as.matrix(read.csv(test_series, sep=";", header=FALSE)) #for ref
+
+       #Simulate serialization in one single call
+       serialize(data_csv, data_bin_file, 350, ";", nbytes, endian)
+       expect_equal(file.info(data_bin_file)$size, 300*50*8+8)
+       for (indices in list(c(1,3,5), 3:13, c(5,20,50), c(75,130:135), 196:200))
        {
-               if (first_write)
-               {
-                       data_line = scan(data, double(), sep=",", nlines=1)
-                       writeBin(data_line, bin_data, size=nbytes)
-                       data_length = length(data_line)
-               }
-               repeat
-               {
-                       data_chunk = scan(data, double(), sep=",", nlines=nb_per_chunk)
-                       if (length(data_chunk)==0)
-                               break
-                       writeBin(data_chunk, bin_data, size=nbytes)
-               }
+               #HACK: as.matrix(as.data.frame( )) required to match (ref) data structure
+               data_lines = as.matrix(as.data.frame( getDataInFile(indices,data_bin_file,nbytes,endian) ))
+               expect_equal(data_lines, data_ascii[indices,])
        }
-       if (first_write)
+       unlink(data_bin_file)
+
+       #...in several calls / chunks of 29 --> 29*10 + 10, incomplete last
+       data_con = file(data_csv, "r")
+       serialize(data_con, data_bin_file, 29, ";", nbytes, endian)
+       expect_equal(file.info(data_bin_file)$size, 300*50*8+8)
+       for (indices in list(c(1,3,5), 3:13, c(5,20,50), c(75,130:135), 196:200))
        {
-               #ecrire file_size-1 / (nbytes*nbWritten) en 0 dans bin_data ! ignored == file_size
-               ignored = seek(bin_data, 0)
-               writeBin(data_length, bin_data, size=8)
+               data_lines = as.matrix(as.data.frame( getDataInFile(indices,data_bin_file,nbytes,endian) ))
+               expect_equal(data_lines, data_ascii[indices,])
        }
-       close(bin_data)
-}
-
-#TODO: read in binary file, always same structure
-getDataFromFile(indices, file, type)
-{
-       bin_data = file(file, "rb")
-       nbytes = ifelse(type=="double",8,4)
-       data_length = readBin(bin_data,"double",1,nbytes) #,endian="little")
-       t(sapply(indices, function(i) readBin(bin_data,"double",n=data_length,size=nbytes)))
-}
+       unlink(data_bin_file)
+       #close(data_con) --> done in serialize()
+})