Help converting cell to matrix of doubles (Large matrix)

Good day Matlabbers!

I have a textfile with almost a million rows that I need to format.

What I want to do is break up a string date into a double matrix that has 3 columns of [yyyy mm dd]. I dont want to make a datevec as I need the months and day.

What I have done is able to extract the date into a new matrix but does anyone have a good suggestion on how to vectorize my code so I dont have to run a loop?

Here is my code

data = {'1991-08-09'  '12:15:00'  66.0  200 '(Unchecked)'  18966.0  200 '(Unchecked).calc';...
    '1991-08-09'  '12:30:00'  66.0  200 '(Unchecked)'  18966.0  200 '(Unchecked).calc';...
    '1991-08-09'  '12:45:00'  66.0  200 '(Unchecked)'  18966.0  200 '(Unchecked).calc'};

% single row
date_short = [str2num(data{1}(1:4)),str2num(data{1}(6:7)),str2num(data{1}(9:10))];

% loop to get all date rows
date_short_loop = [];
for a = 1:3
    date_short_loop(a,1:3) = [str2num(data{a}(1:4)),str2num(data{a}(6:7)),str2num(data{a}(9:10))];
end

Any suggestions are greatly appreciated :) Thanks Norris

Answer 1

[Matthew Eicholtz]

I think datevec is what you want.

d = datevec(data(:,1),'yyyy-mm-dd');
d = d(:,1:3); % if you only want to keep the year, month, and day

Answer 2

[nori lam]

Thanks for the suggestion.

I tried the datevec and it is really slow (comparable to a for loop) which is unfortunate.

I ended up using the find and replace function in Sublime text to remove the '-' marks and just read the whole file with textread.