Pooled mean and standard deviation

Calculate pooled N, mean and STD (using, as inputs, subgroups N, mean and STD)

Calculate pooled n, mean and std from n, mean and std of two groups
(to calculate it to N groups (N>2), repeat it N-1 times)
[npool,meanpool,stdpool] = pooledmeanstd(n1,mean1,std1,n2,mean2,std2)

based on http://www.talkstats.com/showthread.php/7130-standard-deviation-of-multiple-sample-sets

Example:

n1=32;
sample1 = randi(100,n1,1);
mean1= mean(sample1);
std1= std(sample1);
n2=20;
sample2 = randi(100,n2,1);
mean2= mean(sample2);
std2= std(sample2);
n3=9;
sample3 = randi(100,n3,1);
mean3= mean(sample3);
std3= std(sample3);
pool_sample=[sample1;sample2;sample3];
meanpool_real= mean(pool_sample);
stdpool_real= std(pool_sample);

[npooltemp,meanpooltemp,stdpooltemp] = pooledmeanstd(n1,mean1,std1,n2,mean2,std2);
[npool_estimated,meanpool_estimated,stdpool_estimated] = pooledmeanstd(npooltemp,meanpooltemp,stdpooltemp,n3,mean3,std3);

disp(['meanpool_real=',num2str(meanpool_real),' meanpool_estimated=',num2str(meanpool_estimated)])
disp(['stdpool_real=',num2str(stdpool_real),' stdpool_estimated=',num2str(stdpool_estimated)])