Reshape Cell to Matrix

I'm receiving a very large quantity of data in a table like format with only 3 columns (Ticker, Return, Date). What I end up with are thousands of instances of the same date and ticker which becomes very inefficient for Matlab to utilize in calculations. I'd like a matrix with Tickers across the columns and dates down the rows. The trick is that not all tickers have a Return for every date so I can't just reshape the matrix based on a unique count of tickers and dates. Does anyone have any idea how to go about doing this? Here's an example of what I'm looking for.

Initial Format

IBM,1.01,1
IBM,1.04,2
IBM,1.06,3
GOOG,.99,1
GOOG,.98,2
AAPL,1,1
AAPL,1.1,2
AAPL,.99,3

Matrix Format below. Where Dates and Tickers would likely be separate arrays. This example should copy to Excel and make it more clear

,AAPL,GOOG,IBM
1,1,99,1.01
2,1.1,.98,1.04
3,.99,Nan,1.06

Answer 1

[Andrei Bobrov]

EDIT [13:20MSK 23.12.2011]

data = {'IBM',1.01,1
'IBM',1.04,2
'IBM',1.06,3
'GOOG',.99,1
'GOOG',.98,2
'AAPL',1,1
'AAPL',1.1,2
'AAPL',.99,3};

[a,b,b] = unique(data(:,1));
d23 = cell2mat(data(:,2:3));
N = max(d23(:,2));
outdata = accumarray([d23(:,2),b],d23(:,1),[N numel(a)],[],NaN)
out = [{'Name'} a';num2cell([(1:N)',outdata])]

OR

out = [{'Name'},a';...
 num2cell((1:N)'),accumarray([d23(:,2),b],d23(:,1),[N numel(a)],@(x){x},{NaN})];

small corrected in [N numel(a)]

[Brian]

Thanks a lot for the answer Andrei. I think I see where you're attempting to go with this and it might work. I do get an "Index exceeds matrix dimensions" error on the outdata line that I am attempting to debug. If you have a second to look at it today and help me fix the syntax it would be much appreciated.

Thanks,

Brian

[Andrei Bobrov]

Hi Brian! Corrected.

[Brian]

Pure Genius. Extremely useful function that accumarray, I couldn't figure out how to use it properly but I believe I understand it now looking at your revisions. Thanks so much for taking the time to help.

Brian

[Brian]

So I'm having an issue adjusting this function to work with my actual dataset and it may be because I still don't understand the function quite right.

My actual dataset format is as follows....

data = {'06/03/2011','00163T10'

'06/06/2011','00163T10'

'06/07/2011','00163T10'

'06/03/2011','ABCDEFGH'

'06/06/2011','ABCDEFGH'

'06/07/2011','ABCDEFGH'

'06/03/2011','ABCDWXYZ'

'06/06/2011','ABCDWXYZ'};

How would I go about using the arguments above with a dataset in this format. Ultimately, I am attempting to create a logical matrix of whether the ticker was in or out (1 or 0) for each given day.