Bootstrap statistics for multiple column data with varies data points ?

10 views (last 30 days)
aa on 7 Oct 2020
Commented: aa on 8 Oct 2020
Hi everyone,
May someone help me here ...I have data in columns (51 columns)
y = load('Bootstrap_data.txt');
nBoot = 1000;
[bci,bmeans] = bootci(nBoot,{@mean,y},'alpha',.1,'type','per');
bmu = mean(bmeans);
... I want to apply Bootstrap statistics to each column and calcualte mean and its upper and lower bound with 95% confidence interval. I code works well for each colum but this is very laborous to pick column each time and analysis... May some one help me to automate the code and check either it is correct or not ... I placed my code here and attched the data set as well...
It is important to mentioned that few interires in the column are NaN as well.
aa on 8 Oct 2020
nBoot = 1000;
for ii=1:51
[bci(:,ii),bmeans(:,ii)] = bootci(nBoot,{@mean,a(:,ii)},'alpha',.1,'type','per');
bmu(ii,1) = mean(bmeans(:,ii));
I try with this ... but again there is a problem, when i remove the NaN values from, the data ... then the maxtrix dimension no more equal ... in one column we have 51 while in other may be 45 ... so code stop once there is no similar matrix dimension.

Sign in to comment.

Answers (0)


Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!