Importing data from a text file in matrix form

I have a text file (extracted from ANSYS as stiffness matrix)from which I want to obtain a 1710x1710 matrix.Since the text file large,I am attaching it in two parts(as zip files:part-I and part-II).Just copy the second part and paste at the end of first part.Here is part-I.

• The answer should not only work for 1710x1710 case but for all sizes * ( obviously the text format will remain the same )*

[Kamal Bera]

Here is Part-II

[Guillaume]

Can you change the way the file is written so that the index does not collide with the value when the value is negative, that is avoid things like:

5-0.22349156E+04

[Kamal Bera]

Sorry Mr.Guillaume ,the text file is obtained directly through ANSYS. I don't know how to modified as you wanted.But I think it might be separated in excel using text to column option (in excel go to DATA then text to columns).

[Guillaume]

Your text file is basically consisting of fixed width columns. This is a bit more difficult to parse in matlab since there's no built-in function for that. You can either read each line and split them at constant columns or use a regular expression as per Stephen's answer.

[Stephen23]

While regexp is used in my answer, actually a cunning use of sscanf splits that adjacent data. It relies on the fact that it there is either a space or a minus sign...

Answer 1

[Stephen23]

This requires just one small change to the solution to your last question (where the file has the same format):

str = fileread('STIFFNESS MATRIX.txt');

[tok,idx] = regexp(str, 'ROW\s*(\d+)\s+NODE\s*(\d+)\s+[^\n=]+=\s+(\w+)', 'tokens', 'end');
tok = vertcat(tok{:});
idx = [1+idx,numel(str)];

fun = @(b,e)reshape(sscanf(str(b:e),'%f'),2,[]).';
mat = arrayfun(fun, idx(1:end-1), idx(2:end), 'UniformOutput',false);

fun = @(r,v)[sscanf(r,'%i')*ones(size(v,1),1),v];
mat = cellfun(fun, tok(:,1), mat(:), 'UniformOutput',false);
mat = vertcat(mat{:});

out = zeros(size(tok,1));
ind = sub2ind(size(out),mat(:,1),mat(:,2));
out(ind) = mat(:,3);

The only change that I made was to change the regexp token-substring from \s+ to \s*, to allow for the ROW value to occur right next to the ROW text, with or without any space characters. Everything else works just as it did before. The output is too large to be shown, but here are the first and last eight values:

>> out(1,1:8)
ans =
 1.0e+10 *
  0.3212   -4.0406   -0.0000    0.0000   -0.0000   -0.0556   -0.1616    3.2060
>> out(end,end-7:end)
ans =
 1.0e+13 *
       0         0    0.0571    0.0000   -0.0000    0.0000    0.0000    1.0209

which matches the values in the text file. You can check the other 2924084 values somehow...

[Kamal Bera]

It is working.Although I have not checked for other matrix dimensions, hope it would work.Again sorry for any inconvenience caused by me and thank you for your kind consideration. I am accepting your answer.

[Stephen23]

I'm glad to help!