Easiest is to read your file one line at a time, and separate your sections. If your files follow exactly the format you've listed with no missing values, it's simple... Maybe a bit clunky, but serves its purpose. If there's more than height and weight, I'd combine this into a single loop with section logic, rather than create loops for each.
Try this... might contain errors. I just hacked it up here to illustrate.
fid = fopen('data.csv', 'rt');
header = fgetl(fid);
id = regexp(header, ',(.*),', 'tokens');
id = id{1}{1};
% Read height section
fgetl(fid); fgetl(fid); % skip headers
height = [];
while ~feof(fid)
line = fgetl(fid);
vals = sscanf(line, '%f,%f,%f', 3);
if isempty(vals)
break;
end
height(end+1, 1:3) = vals;
end
% Do the same thing for weight.
fgetl(fid); fgetl(fid); % skip headers
weight = [];
while ~feof(fid)
line = fgetl(fid);
vals = sscanf(line, '%f,%f,%f', 3);
if isempty(vals)
break;
end
weight(end+1, 1:3) = vals;
end
fclose(fid);
[EDIT] To select appropriate action from arbitrary order - the dirty way =)
while ~feof(fid)
header = regexp(fgetl(fid), '^(.*),', 'tokens');
if ~isempty(header)
header = header{1}{1}
end
fgetl(fid); % skip second header line
switch lower(header)
case 'height'
% height-processing loop goes here
case 'weight'
% weight-processing loop goes here
otherwise
error('Unexpected section header');
end
end
There are nicer ways to do all this, but my philosophy is if the quick and dirty option is good enough, there's no need to be clever.
