Reading text file

Hi!

This seems like an easy one but I can never wrap my head around reading text files.

I have a text file which contains u and v components of wind and pressure for a cyclone at a certain grid point. Here's a snipet of the data...

 GridID  u wind  v wind  Pressure
 159838   5.25    -1.6  1018.8
 159839   5.19    -2.1  1018.8
 159840   5.03    -2.5  1018.8
 159841   4.76    -3.0  1018.8
 159842   4.42    -3.3  1018.8
#
 160945   5.13    -5.3  1018.5
 160946   4.38    -5.4  1018.6
 160947   3.66    -5.5  1018.7
 160948   3.00    -5.3  1018.8
 160949   2.42    -5.1  1018.8
 160950   1.91    -4.8  1018.9
etc...

The # means that the following data is on another time step. Each timestep is a different size as the cyclone does not affect every grid cell. I want to load this data into a matrix.

Any suggestions on the best way to do this? Any help would be greatly appreciated :)

Answer 1

[Fangjun Jiang]

One way to do it.

clear;
fid=fopen('test.txt');
str=textscan(fid,'%s','headerlines',1);
fclose(fid);
str=str{1};
str=[{'#'};str;{'#'}];
index=find(strcmp(str,'#'));
N=length(index);
Data=cell(N-1,1);
for k=1:N-1
    Data{k}=reshape(str2double(str(index(k)+1:index(k+1)-1)),4,[])';
end

>> Data

Data = 

      [5x4 double]
      [6x4 double]
      [3x4 double]

>> Data{1}
ans =
  1.0e+005 *

     1.598380000000000   0.000052500000000  -0.000016000000000   0.010188000000000
     1.598390000000000   0.000051900000000  -0.000021000000000   0.010188000000000
     1.598400000000000   0.000050300000000  -0.000025000000000   0.010188000000000
     1.598410000000000   0.000047600000000  -0.000030000000000   0.010188000000000
     1.598420000000000   0.000044200000000  -0.000033000000000   0.010188000000000

>> Data{2}
ans =
  1.0e+005 *

     1.609450000000000   0.000051300000000  -0.000053000000000   0.010185000000000
     1.609460000000000   0.000043800000000  -0.000054000000000   0.010186000000000
     1.609470000000000   0.000036600000000  -0.000055000000000   0.010187000000000
     1.609480000000000   0.000030000000000  -0.000053000000000   0.010188000000000
     1.609490000000000   0.000024200000000  -0.000051000000000   0.010188000000000
     1.609500000000000   0.000019100000000  -0.000048000000000   0.010189000000000

>> Data{3}
ans =
  1.0e+005 *

     1.609450000000000   0.000051300000000  -0.000053000000000   0.010185000000000
     1.609460000000000   0.000043800000000  -0.000054000000000   0.010186000000000
     1.609470000000000   0.000036600000000  -0.000055000000000   0.010187000000000

[Leigh]

Thankyou! I used something slightly different but couldn't have done so without this answer!

Answer 2

[Walter Roberson]

We need more information about how you want this stored. You say you want the data stored in a matrix, but do you want all of the data stored in a single matrix, ignoring the '#' boundaries, or do you want one matrix (or at least one cell) for each section ?

If you want everything in one matrix ignoring the '#', then

fid = fopen('YourFile.txt', 'rt');
datacell = textscan('%f%f%f%f', 'HeaderLines', 1, 'CombineOutput', 1, 'CommentStyle', '#');
fclose(fid);
data = datacell{1};

[Leigh]

Cheers, I did want different boundaries, taking into account the # but I've figured it out. Thanks for your reply.