Prev: Paper size not as specified! Print to PDF
Next: Fsolve do not work for my ecuations
From: Matt Fig on 22 Apr 2010 20:20
This might be faster than KRON.


% Data
V = [1 2];
% EXPAND is on the FEX.
F = @(A) [expand(V(:),[size(A,1),1]),A(cumsum(ones(size(A,1),length(V))),:)];
I4 = npermutek(V,4); % Use this to get to npermutek(V,8);

% Engine.
I8n = F(F(F(F(I4))));
I8 = npermutek(V,8);
isequal(I8n,I8) % YES



Note that you could do this much more simply by forming the leading column exactly as I do in the file NPERMUTEK iteslf, instead of trying to use a one line function call. That would probably be fastest of all.
From: Juliette Salexa on 23 Apr 2010 20:47
You're right,
npermutek(1:4,14) is still (much) faster than F(F(I12))

I'll probably have to tolerate the speed for now, and try to learn how your code works later when I get the time,

Thanks for all of the help though!
First  |  Prev  | 
Pages: 1 2
Prev: Paper size not as specified! Print to PDF
Next: Fsolve do not work for my ecuations