Select specific elements in a cell

Hello, this is a very frustrating problem for me

I am using a loop that reads text lines from a data file using fgetl , such a line could look like this (there are many lines):

60 233 24 53 26 53 234 32

And could continue forever. I am trying to create a vector of every 1st value and a separate vector for every 2nd value, which is possible, but the following elements come in pairs . To illustrate:

aline = a b e f e f e f e f e f e f e f ...
anotherline = a b e f e f e f e f e f e f ...
so on ...

I am trying to make a cell for the e and a separate cell for f values, but cannot 'select' them in order to do so. How can you do this? I have tried playing with rem(1:len,2) == 1 (len = length of vector) to select every "odd" or "even" number, but it doesn't work as u want it to, as you cannot do the following.

x = rem(1:25,2) == 0;
vector(x);

What am i missing?

[Image Analyst]

Make it easy for us to help you by attaching your data file.

[dpb]

And be more precise in the explanation of what it is you're selection crietria are--is it the odd or even values by position or is it the value of the content?

You say you cannot do the following

x = rem(1:25,2) == 0; vector(x);

And, I'd say, "why can't you?" It's certainly the hard way to generate the index as

ix=2:2:25;
x=vector(ix);

or simply

x=vector([2:2:25]);

works just fine if you simply fscanf the input.

As IA says, a specific example would help a lot...if it's just a regular space-delimited file then probably the easiest is to read it all in and then select and return the desired columns from memory.

[Kenan Hoyt]

How ironic, i wrote that to make it easy to understand. When i look on it a second time, now it looks more like an assignment that a question, sorry for that :b

Anyway, i solved the problem like this (this is not the whole code, but just to show others who might have the same problem the principle in the solution):

fileLine = fgetl(fid); % gets a line of text
i = 1; % counts each time line 
  while ischar(fileLine)
      numLine = str2num(fileLine); % text line converted to vector
      WW{i} = numLine(1); % define W
      DD{i} = numLine(2); % define D
      for k = 1:floor(length(numLine)-2)/2
          tt{k} = numLine(1+2*k);
          CC{k} = numLine(2+2*k);
      end
  end

And then it reads a new line and i = i+1. The original code was so messy and big that i will spare you the details. Thanks for your time IA-superman and dpb!

Edit: by the way, how do i "solve" this question? I cannot 'accept answer' my own question it seems. Any need to?

[dpb]

for k = 1:floor(length(numLine)-2)/2
  tt{k} = numLine(1+2*k);
  CC{k} = numLine(2+2*k);
end

Is more efficiently written as outlined previously as

tt=numLine(3:2:end);
CC=numLine(4:2:end);

or

[tt CC]=reshape(numLine(3:end),2,[]).';

Indeed, one of the rules of Answers Forum is that can't accept own; that would allow for manipulating the scoring too easily...

[Sean de Wolski]

dpb, you can accept your own answer you just don't get reputation points for it. He can't accept the answer because it's in the comments not the answer box.

[dpb]

Oh, ok...I am now informed on the "ACCEPT" rules--was "underneath the impression" couldn't actually accept one's own.

I added a full solution w/o the line-by-line loop as an additional answer that should solve the problem--if not, hopefully Kenan will come back and explain where there's a problem needing fixup.

Answer 1

[dpb]

Instead of reading the file line-by-line and "growing" the arrays incrementally, try something like

x=textread('filename');
WW=x(:,1);
DD=x(:,2);
tt=x(:,3:2:end);
CC=x(:,4:2:end);