How to get details from header

I have document which contains header and data. One part of header is useless so I skipped this but in the second part are some details which I would like to save as variables.

 the first part of header
  ABR Group Header  ==================================
    Group Number: 2
         Records: 16
  SigGen File #1: D:\data\BioSigData\signaly\Stdpipji33.sig
  SigGen File #2: 
      Subject ID: LEcisty
    Reference #1: usporadani1
    Reference #2: ABR4ch
            Memo: 
      Start Time: Mon Apr 08 11:14
        End Time: Mon Apr 08 11:16
----------------------------------------------------
the second part of header
   Record Number: 24
   Aqu. Duration: 19.9885 ms
     Onset Delay: 0 ms
      No. Points: 488
    No. Averages: 300
   No. Artifacts: 0
      Start Time: Mon Apr 08 11:15

    SigGen Index: 253
       Variables: 
                  Frequency = 7999.98 Hz
                  Attenuation = 0 dB
                  Phase = 180 DegredB
                  Calibration = 0 dB
data
1.18025e-005
1.04013e-005
9.32095e-006
9.15802e-006
9.79093e-006
1.04253e-005
1.02769e-005
9.14981e-006
7.38618e-006
5.5122e-006
3.99292e-006
3.07316e-006
2.70889e-006
.
.
.

I would like save inormation about duration, frequency and number of points. I don't know, how to get this details from header. Regards.

Answer 1

[dpb]

Looks like a fixed format so just count lines and use the textscan ability to be called multiple times on same file...

fid=fopen(filename,'r');
dur=textscan(fid,'Aqu. Duration: %f ms','headerlines',15);
pts=textscan(fid,'No. Points: %f','headerlines',1);
frq=textscan(fid,'Frequency = %f Hz','headerlines',1);
fid=fclose(fid);

May want to cast to variable from cell.

[kubyk]

It works as I wanted. Thank you so much.

Answer 2

[kubyk]

So I get details from head and the data with this code:

 filename = 'usporadani11.txt';
 fid = fopen (filename, 'r');

 dur=textscan(fid,'Aqu. Duration: %f ms','headerlines',14);
 dur=dur{1};
 pts=textscan(fid,'No. Points: %f','headerlines',1);
 pts=pts{1};
 frq=textscan(fid,'Frequency = %f Hz','headerlines',6);
 frq=frq{1};
 att=textscan(fid,'Attenuation = %f dB');
 att=att{1};

 fs=pts/(dur/1000); 

 data=textscan(fid, '%f', 'headerlines', 2);
 data=data{1};

 fclose(fid);

But I can have several this sets in one file. It looks: header1, data1, header2, data2...header16, data16. Still the same format. Is it possible to get this ? Regards

[dpb]

If it's fixed length or you can count the lines based on input in the headers, just keep on doing the same thing in a loop.

You can allocate additional cell arrays and concatenate the results or use named dynamic fields in a structure or convert the cell contents of the textscan call to an array and concatenate as storage mechanism depending on needs.

OBTW, you can save a step on the conversion--instead of

dur=textscan(fid,'Aqu. Duration: %f ms','headerlines',14);
dur=dur{1};

can write

dur=cell2mat(textscan(fid,'Aqu. Duration: %f ms','headerlines',14));

Would be nice to have the facility to return variables instead of only cells from textscan, but not to be... :(