write output to a .txt or .mat file

Hi everyone!

I have a couple of huge data files in .txt format and I´m creating 15-min intervals for these files so that each huge file gets reduced to a size of 34x42. The data files have the following structure:

datetime           name    price1    price2    price3 up to price40
19.10.09 09:00:00  basf       30       35       33    .....
19.10.09 09:15:00  basf       30       35       33    .....
19.10.09 09:30:00  basf       30       35       33    .....
19.10.09 09:45:00  basf       30       35       33    .....
...
19.10.09 17:30:00  basf       30       35       33    .....
My first question is how can I save such a file structure into a .txt or .mat file? because date and name are strings, and all prices are doubles. How can I put all this together? I thought of using a cell array but Im not sure because I´m really new to matlab and I´m learning it by doing. Also I tried:

save myvariables.txt datetime name price1 price2 price3

but it wont work because datetime and name are strings a nd the prices are doubles.

I also tried this:

save myvariables.txt price1 price2 price3

but the problem is that when i load it, the variables are loaded into my workspace as single variables and I can´t see thefile as the structure I mentioned above. How can i do it then?

So my second question is, , how could I save all these reduced files into a single .txt file or .mat file?? what kind of loop can I use to append all these reduced files into one single file?

I would really appreciate any help because I don´t know what else to do !

Thank you so much for your help :) Have a nice week!

Lourdes

Answer 1

[bym]

How about

fprintf()
dlmwrite()

[Lu]

Dear proecsm,

Thank you for your answer! Im trying it but still no sucess yet. Ill keep trying.

Have a nice day :)

Answer 2

[Ivan van der Kroon]

If you save and reload your data with

y=load();

you have a structure file anyhow. If you want to save your variables in time1.mat

save('time1.mat','datetime','name','price*');

(the * after price will save all the variables that start with price in their name, the so-called wildcard). If you now type

time1=load('time1.mat')

time1 = 

      price1: 30
      price2: 35
      price3: 33
        name: 'basf'
    datetime: '19.10.09 09:00:00'

Is this what your are looking for? If you might wonder how to implement this in a for-loop such that the number in 'time#.mat' goes up, you can use eval, as in

eval(['save(''time' num2str(k) '.mat'',''datetime'',''name'',''price*'');'])

where k is the number you want to assign to the mat-file. The double '' are to hold them inside the string class which eval requires in this case.

Furthermore, I would suggest to concatenate the prices in a single variable, i.e. price=[price1 price2 price3... ]. Just suggesting...

[Lu]

Dear Ivan,

Thank you for the reply! My script reduces huge data files into an output matrix of size 34x42. The problem is that I have around 205 and i can´t run the script for each file individually. So I created a for loop so it would do the calculations for all 205 files, and it would store the variables I need in a matrix called Output for each file. I just need to append all these output matrices one below the other one and so on, so i can just get 1 final matrix of size 6970x42.

I guess i will open a new thread and explain it better there.

Thank you so much for your help though :)

Answer 3

[Ivan van der Kroon]

m=34;
n=42;
N=205;
A=zeros(m*N,42);

for j=1:N
%load matrix B, don't know how you saved it, so just assuming you do it here and have variable B of size 34x42.
A((j-1)*n+(1:n),:)=B;
end