Having difficulty reading in a file with csvread()

Hi All,

I have a text file which looks like this:

Date,Col1,Col2
2012/06/06 11:40:12,1.01849,1.01881
2012/06/06 11:40:13,1.0185,1.01881
etc

I want to read this into a matrix like this:

[2012 06 06 11 40 12 1.01849 1.01881;
 2012 06 06 11 40 12 1.01850 1.01881
etc
]

I can't seem to do this with csvread or textscan.

Any ideas?

Answer 1

[dpb]

csvread can't because it's not comma-delimited all fields so that's a non-starter

Air code--

fmt='%d/d/d d:d:d,%f,%f';
d=cell2mat(textscan(fid,fmt,'headerlines',1,'delimiter',''));

Caveat: As said, untested and while I think the 'delimiter','' and explicit delimiter matching will work, the C input parsing often throws a curveball. Anyway, that's where I'd start.

Answer 2

[Stewart Charles]

Excellent. Thanks for your help. The actual code turned out to be:

c = textscan(fid, '%d/%d/%d %d:%d:%d,%f,%f','HeaderLines',1,'Delimiter',' ');
d = cell2mat(c(1:6));
v = cell2mat(c(7:8));

Note: I added some % and also found that you can't have ints and floats in the one matrix.

Thanks again!

[dpb]

Yeah, don't know why didn't get all the % signs in there--I guess 'cuz I'm a Fortran guy at heart and detest the C format strings, maybe... :)

I guess I hadn't ever noticed that problem before -- but it's simple enough to work around as one doesn't have to use %d for the integer values...

>> type stew.txt
Date,Col1,Col2
2012/06/06 11:40:12,1.01849,1.01881
2012/06/06 11:40:13,1.0185,1.01881

>> d=cell2mat(textscan(fid,'%f/%f/%f %f:%f:%f,%f, %f', ...
              'headerlines',1,'collectoutput',1));
>> num2str(d,'%.4f')
ans =
2012.0000   6.0000   6.0000  11.0000  40.0000  12.0000   1.0185   1.0188
2012.0000   6.0000   6.0000  11.0000  40.0000  13.0000   1.0185   1.0188
>>

Answer 3

[Image Analyst]

Wow. Complicated. Hopefully you have R2013b or later because by far the easiest way is to just simply create a table with readtable():

t = readtable('test.dat')