Find the Argmin of a vector

Help:

Let x be a vector and B_1...B_p p matrices,

I look for the index j* of the matrix such that

||B_j*||=min ||B_i x||, (||.||=norm)

How do I write this with matlab?

[Matt J]

Which norm?

[Mohsen Dlala]

Here,all norms are equivalent, euclidian for example

[Walter Roberson]

http://www.mathworks.com/matlabcentral/answers/304528-tutorial-why-variables-should-not-be-named-dynamically-eval

[Mohsen Dlala]

@Walter Roberson

As I'm a beginner in Matlab, I didn't understand anything, anyway, thank you very much for your answer.

[Walter Roberson]

You have variables named B_ followed by a number, and you want to process the whole series of variables. That is bad programming practice. You should have avoided naming the variables that way, and should have used (for example) a cell array.

Are you trying to find which of the B_ arrays leads to the smallest norm given a vector x? If so then if you had the arrays represented as a cell array, that would only be a small number of statements.

Answer 1

[Matt J]

You should have avoided naming the variables that way, and should have used (for example) a cell array.

Indeed. Here's an example of how to do it when that is the case.

Bcell={rand(5), rand(5), rand(5), rand(5)}; %B1,...,Bp   p=4

[~, jstar] = min(cellfun(@(B) norm(B*x),Bcell)) %EDITED

[Walter Roberson]

I think it would be more

[~, jstar] = min(cellfun(@(B) norm(B*x),Bcell))

[Mohsen Dlala]

By applying the two Bcell propositions you gave, Compiling several times, the jstar value changes, and you don't get the correct jstar value.

We notice that there is a random choice

[Matt J]

The result of the code is random because the Bcell{i} in the example were randomly generated according to rand(5). You were meant to insert your own matrices.

[Mohsen Dlala]

Thanks very mach Matt

[Matt J]

You're very welcome, but please Accept-click the answer if you got it working.