Question

I have a huge data, with 7 colums and 20000 rows. I let Matlab to read mydata:

[tdata zdata kdata ldata mdata ndata pdata]=textread('mydata.txt')

But what I need is something else. I want to let Matlab to collect every 1000. Row from the data. Help to textread give an explanation for the first nth data:

[tdata zdata kdata ldata mdata ndata pdata]=textread('mydata.txt',n).

Is there any way to do this with a small change of textread format? Or should I write a for loop?

Answer 1

It’s probably easiest if you read all data, and crop away the unwanted data afterwards:

[tdata zdata kdata ldata mdata ndata pdata] = textread('mydata.txt')

tdata = tdata(1:1000:end);
zdata = zdata(1:1000:end);
kdata = kdata(1:1000:end);
ldata = ldata(1:1000:end);
mdata = mdata(1:1000:end);
ndata = ndata(1:1000:end);
pdata = pdata(1:1000:end);

If the memory overhead is too large, or you find this unacceptable, you’ll have to write a loop with fgetl, something along these lines:

fid = fopen('mydata.txt', 'r');
i = 0;
while (~feof(fid))
    i = i + 1;
    line = fgetl(fid);

    if mod(i, 1000) == 0
        parsed = textscan(line, '%f%f%f%f%f%f');
        %# etc.

    end
end