how to import large data file into matlab

I have very large data 5GB space delimited mixed data (data.prn) with column header:

                         *0000*
                         *0000*
       al      10.598010201     1.890025647
       am      11.222025789    -5.369021654
       as     -10.256862254    11.235854322
       at      14.589056356    15.888792312 
               .           .
               .           .

I know it has three column but i don't know exactly the number of rows since the data is to large to open. My target is to plot data(:,1) vs data(:,2). Any suggestion which command is fast to import the data to matlab.Thanks!

Answer 1

[Walter Roberson]

fid = fopen('data.prn', 'rt');
datacell = textscan(fid, '%*s%f%f', 'HeaderLines', 2);
fclose(fid);
plot(datacell{1}, datacell{2})

However, keep in mind that your screen only has a few thousand pixels, not billions, so you are going to end up with multiple data points per pixel, which will be slow and of no good visual purpose. You should consider plotting only a subset of the data, such as

plot(datacell{1}(1:1000:end}, datacell{2}(1:1000:end))

[Abebe kebede]

Thanks Walter.

[Abebe kebede]

@Walter, It seems the output is no as i want it. When i used the above code and try to read datacell{:}i get

10.598010201

NaN

10.598010201 ... and so on

but i want the result to be like

        [10.598010201]     [1.890025647]
        [11.222025789]     [-5.369021654]
        [-10.256862254]    [11.235854322]
       and so on. Thanks