In your example data, each value is represented with two bytes: one for the value, one for a comma or newline. Each Int64 value is 8 bytes. If all your values are between 0 and 255, you could use UInt8 to represent them and cut the size in half.
On Tue, Jul 21, 2015 at 1:16 PM, paul analyst <paul.anal...@mail.com> wrote: > I have data in txt file, some milons like this: > 0,0,2,0,0,0,0,0,0,0,0,0,0,0,0,0,0 > 0,0,0,0,0,0,0,2,0,0,0,2,0,0,0,0,1 > 0,0,0,0,1,0,0,0,1,0,0,0,0,0,0,0,1 > > Coding win1250. > > size of dane.txt is 1.3 GB > > D=readcsv("dane.txt") > k,l=size(D) > > using HDF5, JLD > hfi=h5open("D.h5","w") > close(hfi) > > fid = h5open("D.h5","r+") > g = fid["/"] > dset1 = d_create(g, "/D", datatype(Int64), dataspace(k,l)) > dset1[:,:]=D > close(fid) > > After save to h5 file the file has 6.3 GB ? Why new file is 4 times biger? > Paul >