How to best scan a table for same element values and create arrays to store the extracted data successively?

 I have a dataset like such: A text file with three columns Var1, Var2 and Var3.

 In Var1 there are many elements of the same name to be exact in the following format: 0000000004 and when it reaches the 100 it becomes 0000000100.

 UPDATE: I forgot to mention I used readtable to read in the values and set them up in a table in the below format.

 My goal is to loop through the elements of Var1 and pick out the Var2 and Var3 values of the same Var1 and save them to a seperate array holding Var2 and Var3

 Example:         Var1         Var2          Var3
            0000000004         5             6
            0000000004         7             8
            0000000004         9             10
            0000000005         11            12
            0000000005         13            14
                ..             ..            ..

 This will become Array4 =    4     6
                              7     8
                              9     10
                  Array5 =    11    12
                              13    14
and so on..

 being new to matlab(coming from C++ background) the task has been harder than I anticipated. The place where I am struggling is to write a loop that changes the values I am searching for automatically, like how do I replace the numbering from 0000000004.png to 0000000010.png (replacing the zero with a 1 without going through a lot of trouble) and I am not sure how to search for a string value (since it is not only numbers). Could you please give me a few hints of how to start ?

- I am going to iterate through the values of variable one via a for loop 
- when there is a new value of Var1 I will create an array for storing its Var2 and Var3 values 
- the whole procedure is put in a for loop to make it easier for the process to be repeated when a new 000000000x.png value appears in the Var1 column

Answer 1

[Guillaume]

1) Do not create numbered variable names (names that you would have to generate dynamically), use the adequate container for your data, which is a cell array in your case.

2) You mention a table in your question's title but a file in the question's body. So what have you got?

3) I would read the file all at once in a cell array and then use unique and a loop to split the cell array into what you want:

fid = fopen('somefile.txt', 'rt');
allfiledata = textscan(fid, '%s %f %f'); %create a cell array with 3 columns

[var1, ~, locations] = unique(allfiledata{1});
subarrays = cell(1, numel(var1));
fullmatrix = cell2mat(allfiledata(2:end));
for varindex = 1:numel(var1)
  subarrays{varindex} = fullmatrix(locations == varindex, :);
end
celldisp(subarrays)

[KBK007]

Thank you for your reply, I meant i used readtable to read in the data as they are grouped in the text file as follows:

on each line there is 000000000x.png doublevalue doublevalue

the ultimate goal is to compare the value of for example

 0000000004.png    doublevalue    doublevalue

from the first file and see if there exists the same data in the other file, that is why I want to store the values as such.

Anyways my code so far is sorta like the following:

 file1 = readtable('file1.txt',...
    'Delimiter',' ','ReadVariableNames',false);
 file2 = readtable('file2.txt',...
'Delimiter',' ','ReadVariableNames',false);

 for m = 4 : 163
    for i = 1 : size(DetectedCorners.Var1)

        strcmp(DetectedCorners.Var1(i),'000000000.png'); %%I want to reference m here so it would change dynamically during the comparison. 
        %%do what you posted above

     end
 end

[KBK007]

point out

m = 4 : 163 because I have 00000000m.png images and the numbering starts at 000000004.png and ends at 000000163.png

[Guillaume]

Code gets reused. In the future your file numbers may be outside the range [4, 163] so I wouldn't hardcode it particularly as I've shown you can easily extract what files you have exactly. So you don't waste time comparing filenames that may not be present and you don't miss filenames that actually are present.

With a table, it's simply:

[filenames1, ~, rows1] = unique(file1.var1);
[filenames2, ~, rows2] = unique(file2.var1);

However, if what you want to find if which data is common to both file, there's a command for that:

commondata = intersect(file1, file2);

[Image Analyst]

If your numbers are double with fractional parts, rather than integers, see the FAQ.

[KBK007]

@Guillaume Thank you for your answer, I get it now. The problem is going from C++ to Matlab is not an easy process, I have to get used to how Matlab works with its data. That is why I am reading a guide and acquiring the logic behind data handling in Matlab.