Writing rows/columns from .csv to matrix using Matlab

Question

Using a .csv file, I am attempting to read a set of rows from a .csv file and then write each row to a matrix.

An error occurs that states:

 'Index exceeds matrix dimensions.'
 Error in Id_Vg (line 17)
        meas(j,:) = csvread(filename,0,0,[row 2 row+j 5]);

Is the issue that this line is attempting to write columns 2 to 10 into one element of 'meas'?

I thought that preallocating 'meas' with zeros would allow each individual .csv column and row to be written to 'meas'.

function [ output_args ] = Id_Vg(filename, row,...
                             nopoints, noskip);

%directory= address of the .csv files "C:\....."
%filename = .csv name "1us.csv"
%row      = starting row to ignore all text before values, "120"
%nopoints = number of measurement points at each condition "[1e1 1e3 1e1 1e3 1e2]"
%noskip   = determines whether to skip condition "[0 1 0 1 0]" 0=skip.

rowstop = row+sum(nopoints);
M = csvread(filename,row,1,[row 2 rowstop 5]);

meas = zeros(sum(nopoints),4); 

    for i = 1:length(nopoints)
        if noskip(i) == 0
            for j=1:sum(nopoints)
        meas(j,:) = csvread(filename,0,0,[row 2 row+j 5]);
            end
        end
    end

Edit: Changed 'meas(j)' to 'meas(j,:)' and altered the useful .csv columns so that no empty columns are written to 'meas'.

Error still occurs:

Subscripted assignment dimension mismatch.

I understand this is usually due to, in this case, writing too many .csv columns to a matrix that does not have sufficient columns. This should not be the case as even writing a higher number of zeros to meas produces the same error.

if csvread read 10 value on each row then you should write: meas(j,:) = .... — obchardon
– obchardon, Commented Jun 2, 2016 at 12:36
The same error occurs but thanks. This should help with writing the .csv columns to 'meas'. Possibly it's a different issue to what I mentioned in the original comment. — Azaxa
– Azaxa, Commented Jun 2, 2016 at 12:49

obchardon · Accepted Answer · 2016-06-02 12:50:10Z

1

You're right: the issue is that this line is attempting to write columns 2 to 10 into one element of 'meas'.

You should write meas(j,:) = ... instead of meas(j) = ...

meas(j,:) is equivalent to meas(1,1:10).

if you write meas(j) = [1 2], you're trying to write 2 element into one element of meas which is impossible.

And be carefull with this code your csvread should always return 10 values !

answered Jun 2, 2016 at 12:50

obchardon

10.8k1 gold badge19 silver badges36 bronze badges

Sign up to request clarification or add additional context in comments.

5 Comments

Azaxa Over a year ago

Unfortunately it appears there is a dimension mismatch even when considering your last statement. The post has been editted with your recommendations, thanks.

Azaxa Over a year ago

Obchardon, is there anything more you could suggest?

obchardon Over a year ago

I also think that your parameter are wrong in csvread: because when you write: csvread(filename,0,0,[row 2 row+j 5]); it means that you get some value on more than one row ! Shouldn't It Be csvread(filename,0,0,[j 2 j 5]);` ? So you read 4 values of only ONE row ! Or csvread(filename,0,0,[row+j 2 row+j 5]);

Azaxa Over a year ago

That is very true! I missed that one, thanks. Still, unfortunately the "Subscripted assignment dimension mismatch" error still persists...

obchardon Over a year ago

Now your code is correct but I don't know how your data are structured, I cannot help you.

Collectives™ on Stack Overflow

Writing rows/columns from .csv to matrix using Matlab

1 Answer 1

5 Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

5 Comments

Your Answer

Sign up or log in

Post as a guest

Related