How to generate a random matrix with a constraint or step range ?

Question

Varun Pai 2020-4-23

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/520058-how-to-generate-a-random-matrix-with-a-constraint-or-step-range

评论： Varun Pai 2020-4-23

I want to generate a Nx4 random matrix. In this matrix, each column is considered having different range and step. The random values should be only from these steps.

For e.g. 1st column 0:0.1:1

2nd column > 0:1:10

3rd column > 2:0.5:8

4th column > 0:-1:-10

I want to generate a random matrix of N=5. I should get a matrix with above conditions. Like this

[0 5 3.5 -5 ;

0.8 8 7 -3

0.2 0 2.5 -7

0.6 8 6.5 0

0.9 3 7.5 -9]

The N value could go to any size. Repeated values are permissible. I tried some random functions like randperm, randsample etc but I am not able to produce like this. Randperm is not allowing to produce repeated values.

How could I generate it ? Could you help with a code sample ?

0 个评论
显示 -2更早的评论隐藏 -2更早的评论

请先登录，再进行评论。

请先登录，再回答此问题。

Answer 1

KSSV 2020-4-23

0
链接

此回答的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/520058-how-to-generate-a-random-matrix-with-a-constraint-or-step-range#answer_427802

编辑：KSSV 2020-4-23

在 MATLAB Online 中打开

Read about randsample, randperms.

C1 =  0:0.1:1 ; 
C2 = 0:1:10 ; 
C3 = 2:0.5:8 ; 
C4 = 0:-1:-10 ;
N = 5 ; M = 4 ; 
C = zeros(N,M) ; 
C(:,1) = randsample(C1',5) ; 
C(:,2) = randsample(C2',5) ; 
C(:,3) = randsample(C3',5) ; 
C(:,4) = randsample(C4',5) ; 

3 个评论
显示 1更早的评论隐藏 1更早的评论

KSSV 2020-4-23

编辑：KSSV 2020-4-23

You split the big number into multiples and pick from each column.

You can use repmat on C1, repeat the values and pick them at once.

Varun Pai 2020-4-23

在 MATLAB Online 中打开

That's great... Thank you for the suggestion.

function ParameterSet = GenerateRandomSet(MinVal,MaxVal,StepVal,N)
% MinVal : 1 x M array of Lower limit
% MaxVal : 1 x M array of Upper limit
% StepVal : 1 x M array of Step values
% N : ParameterSet row size
% Usage eg. GenerateRandomSet([0 0 2 0],[1 10 8 -10],[0.1 1 0.5 -1],50)
rows = N;
cols = length(MinVal);
ParameterSet = zeros(rows,cols);
for c = 1:cols
    ParamVectors = MinVal(c):StepVal(c):MaxVal(c);
    repfactor = floor(N/length(ParamVectors)) + 1;
    ParamVectors = repmat(ParamVectors,1,repfactor);
    
    ParameterSet(:,c) = randsample(ParamVectors',N); 
end
end