How to import .dat files (comprising multiple columns which can only be copied and pasted all at the same time) into MATLAB?

I'm currently trying to import a .dat file (e.g.0002_03.dat) into MATLAB in order to logically arrange the data in a spreadsheet.

My issue is that, although the data in 0002_03.dat is nicely ordered into columns separated by spaces, it is impossible to select, copy and paste one column at a time (which would make my question redundant since, in that case, I wouldn't have to import the data into MATLAB in the first place - I could copy and paste each column into columns in Excel). If I try to select one column only, I end up selecting the entire dataset.

Thus far I have tried: import the .dat file in MATLAB using 'import' - the result is a 51x1 cell array; fid=fopen('0002_03.dat','r') which returned -1; textscan('0002_03','.xls') which returned 'Empty cell array: 1-by-0'

[Ingrid]

just a small note, not matlab related. You can always just open the dat file in excel directly if that is what you want to achieve. Just right click on the file, select open with. Normally the data is automatically ordered correctly in excel, one value per column and you can select the columns you want. If this is not the case, you can always use the "text to columns" option (which you can find in the data ribbon) and select the correct delimiter

if you want to use matlab because you will use it in your further analysis you should have a look at textscan

doc textscan

[Bianca Elena Ivanof]

Hey Ingrid,

Many, many thanks for your quick answer - the 2nd option seems to be working, but for some reason it only converts the text into columns in some cases only. For example, columns 16, 17, 18, 19, 20 don't seem to be treated as separate columns when clicking on the 'text to columns' option.

Unrelated to our discussion, my supervisor (who did this in Excel) mentioned that there might be an even quicker way of basically telling MATLAB to treat the data in the cell array as a normal nxm matrix. But I am a bit skeptical, since the fid=fopen('filename') function returned -1, which I know that means that the file cannot be open period.

Again, thank you for your solution anyway.

[Ingrid]

your supervisor is right, MATLAB can do that very easily and fast

if your fid == -1 something else must have gone wrong because you are able to open the file in Excel. Are you sure that you provide the full path of the filename, together with the extension?

[Jan]

There is a reason for FOPEN to fail. Most likely a typo in the file name or you are in the wrong folder.

[Bianca Elena Ivanof]

I've set the path datapath = 'D:\foldername\subfolder1\subfolder2\etc.' fopen(fullfile(datapath,'filename'), 'r')

ans: fid=3...

I've also tried fopen(fullfile('D:\folder\subfolder\etc.','filename'), 'r') (which is pointless since i've created a variable called datapath containing the path of the .dat file; but I've said maybe the fopen and fullfile functions will give me what I want if I rewrite the whole path again)

ans: Error: Unexpected MATLAB operator.

Does any of you guys have any idea what's going on? I have the hunch what I'm trying to achieve (see main question) is so simple, yet I can't seem to get my head around it at the moment.

Many thanks in advance

[Jan]

Please post the complete error message.

Answer 1

[Jan]

Are you in the correct folder?

DataPath = 'C:\Temp';  % Adjust accordingly!
fid      = fopen(fullfile(DataPath, '0002_03.dat'), 'r')

[Bianca Elena Ivanof]

Hey Jan,

Firstly, many thanks.

Set the path and wrote fid = fopen(fullfile(DataPath, '0002_03.dat'), 'r')

Got fid = 3, which confused me even further...

[Jan]

fid = 3 means, that Matlab has opened the file successfully. What does confuse you at this point? I think, this is a valid point to start from.

What was the problem with the data imported by the import function? I still do not understand, which problem you are struggeling with.