data generation using bootstrap method

 Hi all,
 I have a data series like 'a':

                         a= [50;46;28;31;38;49;52;33;38;32;61;90;42;42;45;29;36;30;27;20;31;30;34;23;]; 

 I want to genarate more data which originated from 'a' using bootstrap method.

 How can I write this program?
 Thanks in advance.

Answer 1

[José-Luis]

Bootstrapping consists in selecting a subset of the data. That sounds like a job for randperm()

 a = randi(60,1,50);
 subsetSize = numel(a) - 1; %How many values in your bootstrap
 numSamples = 10;  %number of simulations
 your_data = ones(numSamples,subsetSize); %one sample per row
 for ii = 1:numSamples
     your_data(ii,:) = a(randperm(subsetSize));
 end

Note that you could also generate all possible subsets:

 allSets = nchoosek(1:numel(a),subsetSize);
 your_data = a(allSets);

And select your data from it, if you want to avoid repeated values.

[peach_R]

I wonder why it starts from a = randi(60,1,50) as I think that 60 means integer from 1 to 60 not from elements in 'a'.