MATLAB Answers

Calculate standard deviation for different groups

20 views (last 30 days)
leonidas86
leonidas86 on 23 Jul 2018
Commented: Guillaume on 23 Jul 2018
I want to plot a Measurement with a bar figure. Therefore I want to create different groups and calculate the standard deviation for each group. I have a vector "Measurement" which is sorted according to increasing values. I get group IDs for the Measurement values by the use of the histc function.
[~,Measure_ID]=histc(Measurement_x,0:200:max);
In the next step I try to calculate the standard deviation with the accumarray function.
group_std = accumarray(Measure_ID,Measurement_y,[],@std);
The std command uses the follow mathematical formula:
The problem is that accumarray calculates the arithmetic average for each group separately. But I want to see the standard deviation regarding to the global arithmetic average. Is there a way to solve my issue?
Version: 2012b
  1 Comment
Jan
Jan on 23 Jul 2018
Do you mean:
std(Measurement_y)
? What exactly is "standard deviation regarding to the global arithmetic average"?

Sign in to comment.

Accepted Answer

Jan
Jan on 23 Jul 2018
Edited: Jan on 23 Jul 2018
If you want to set the x_bar values of the formula to mean(Measurement_y), why not implementing this as function?
function s = specialStd(x, m)
s = sqrt(sum((x - m) .^ 2) / (length(x) - 1));
end
and call it like:
m = mean(Measurement_x);
group_std = accumarray(Measure_ID, Measurement_y, [], @(x) specialStd(x, m));
  1 Comment
Guillaume
Guillaume on 23 Jul 2018
specialStd, notexactlystd, I think we can agree that this thing shouldn't be called plain std!

Sign in to comment.

More Answers (1)

Guillaume
Guillaume on 23 Jul 2018
"The problem is that accumarray calculates the arithmetic average for each group separately".
It's not accumarray doing that, it's std, because that the definition of the standard eviation.
To do what you want, you'll have to implement your own standard deviation calculation:
whole_mean = mean(Mesurement_y);
group_notexactlystd = accumarray(Measure_ID, Measurement_y, [], @(v) sum((v - whole_mean).^2)/(numel(v)-1));

Community Treasure Hunt

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

Start Hunting!