Genetic Algorithm - Vectorized Mode - Reg

'Position' is a (68,6) array of doubles representing the population of Genetic algorithm. 'classGA1 ' is a (68,1) cell array of the type of population class ( inventory class A,B or C) . 'ClassDM1 ' is a (68,1) cell array of the inventory class(A,B, or C)given by the decision maker (materials manager).

[Fitness] = InvClassifyGAFitnessFunc(Position,classGA1,ClassDM1); 

is the Fitness is a (68,1) double array determining the fitness of population.

I want to operate the GA in vectorized mode.

[Fitness] = @(Position)InvClassifyGAFitnessFunc(Position(':',1:6),classGA1{':',1},ClassDM1{':',1});
% vff = @(Position) InvClassifyGAFitnessFunc(Position(':',1:6), classGA1{':'},CDM1{':'});
 A=[0,0,0,0,-1,1];
 b=[0];
 Aeq=[1,1,1,1,0,0];
 beq=[1];
 lb=[0,0,0,0,0,0];
 ub=[1,1,1,1,1,1];
 options= gaoptimset('PlotFcn',@gaplotbestf,'Vectorized','on');
 [x,fval]=ga(Fitness,6,A,b,Aeq,beq,lb,ub,[],options);
 
 

The genetic algorithm is giving the error message

Error using InvClassifyGAFitnessFunc

Too many input arguments.

How to run my genetic algorithm in vectorized mode.

Answer 1

[Walter Roberson]

'classGA1 ' is a (68,1) cell array of the type of population class

When that is the case, then

classGA1{':',1}

expands into 68 different arguments, equivalent to classGA1{:}

[madhan ravi]

+1 sir Walter really humbled(gentled) with your answer as always, interesting was the fact is classGA1{':',1} (was really new) is equivalent to classGA1{:} (which I was aware of this).

out of curiosity : Is there any link about this notation so that more can be learn't?

[Walter Roberson]

This gets down to some some internals on how MATLAB processes the : operator . When : is used by itself as a parameter, then what gets passed is ':' (the character), and subsref detects that and understands it as requesting the entire array along that dimension.

https://www.mathworks.com/help/matlab/ref/subsref.html

[Stephen23]

@madhan ravi: indexing using parentheses, curly braces, etc, is just a convenience notation for subsref. You can see examples in its help:

https://www.mathworks.com/help/matlab/ref/subsref.html

And also some threads on this forum, showing how this is useful:

https://www.mathworks.com/matlabcentral/answers/422001-changing-dimension-in-a-loop

https://www.mathworks.com/matlabcentral/answers/367188-how-to-filter-an-n-d-array-with-matrix-indexing

https://www.mathworks.com/matlabcentral/answers/344486-generate-comma-separated-list-in-single-line-of-code

This is also why those tutors who give assignments which boldly state "inbuilt functions cannot be used" really are making life very hard for their pupils!

[madhan ravi]

Thank you!

Answer 2

[madhan ravi]

Your function requires only two inputs whereas you have stuffed in 3 .

[Siva]

shall i have to combine the functions 'classGA1' and 'ClassDM1' into a single function ( say name of the combined function is 'class_new' ) .Then it will be o.k?

[madhan ravi]

I don't have any experience with GA but seeing your previous question but it's clear that the function was made with two input arguments there.

[Siva]

i combined the two input arguments to the fitness function into a single argument and ran the GA. but the same error message too many input arguments in the fitness function is coming.

[Stephen23]

"Your function requires only two inputs whereas you have stuffed in 3 ."

There are actually 137 inputs to InvClassifyGAFitnessFunc:

https://www.mathworks.com/help/matlab/matlab_prog/comma-separated-lists.html

Three inputs is probably the correct number, judging by this earlier thread:

https://www.mathworks.com/matlabcentral/answers/435093-genetic-algorithm-reg#answer_351648