How to convert a text file into a table properly

Hi,

I have text file that looks like this:

time-->data1-->data2-->data3-->...

0-->1.2,2.9,3.4-->4.6,5.5,3.2-->0.1,0.1,0.9-->...

0.1-->2.2,2.4,3.4-->4.8,5.8,3.8-->0.1,0.2,0.9-->...

...

Here, --> represents a tab.

I would like to create a table out of it so that the table look like this:

0 {1.2,2.9,3.4} {4.6,5.5,3.2} {0.1,0.1,0.9} ...

0.1 {2.2,2.4,3.4} {4.8,5.8,3.8} {0.1,0.2,0.9} ...

...

To do so, I first use readtable:

T_data = readtable('my_data.txt', 'Delimiter', '\t');

This will the create a table like this:

0 {'1.2,2.9,3.4'} {'4.6,5.5,3.2'} {'0.1,0.1,0.9'} ...

0.1 {'2.2,2.4,3.4'} {'4.8,5.8,3.8'} {'0.1,0.2,0.9'} ...

...

As you can see, each element of the table is a cell containing a string. Now I use a for loop with textscan to change the string cells into numeric cells:

for row = 1:height(T_data)
    for col = 2:width(T_data)
        T_data{row, col} = textscan(T_data{row, col}{1}, '%f', 'Delimiter', ',');
    end
end

And it does the trick, however, the second part is very slow! (A 4200*8 table took around 10 seconds.)

I was wondering if there are alternative solutions that are more eficient.

Thank you in advance.

Answer 1

[Chunru]

T = readtable('test.txt', 'Delimiter', '-->')

T = 2×4 table
    time         data1              data2              data3     
    ____    _______________    _______________    _______________

      0     {'1.2,2.9,3.4'}    {'4.6,5.5,3.2'}    {'0.1,0.1,0.9'}
    0.1     {'2.2,2.4,3.4'}    {'4.8,5.8,3.8'}    {'0.1,0.2,0.9'}

for i=1:size(T,1)
    for j=2:size(T,2)
        T{i, j}{1} = str2num(T{i, j}{1});
    end
end
T

T = 2×4 table
    time             data1                       data2                       data3          
    ____    ________________________    ________________________    ________________________

      0     {[1.2000 2.9000 3.4000]}    {[4.6000 5.5000 3.2000]}    {[0.1000 0.1000 0.9000]}
    0.1     {[2.2000 2.4000 3.4000]}    {[4.8000 5.8000 3.8000]}    {[0.1000 0.2000 0.9000]}

[Shayan Taheri]

Thank you. Yes, I have also tried this solution. But the execution time is even longer for some reason!

Answer 2

[Chunru]

Try this:

fid = fopen('test.txt', 'r');
fgetl(fid);   % skip the header
i=1;
x = [];
while ~feof(fid)
    s = fgetl(fid);
    tmp = textscan(s, '%f', 'Delimiter', {',', '-->'});
    x = [x; tmp{1}'];
end
fclose(fid);
x

x = 2×10

0 1.2000 2.9000 3.4000 4.6000 5.5000 3.2000 0.1000 0.1000 0.9000 0.1000 2.2000 2.4000 3.4000 4.8000 5.8000 3.8000 0.1000 0.2000 0.9000