Find range of values where 80% lies

Hi,

I've got gradually changing data set in an array, and I'm trying to find the threshold where 80% of the data lies below and hopefully get the corresponding indices. A Gaussian distribution takes it about the mean but using this will also most definitely remove the "good" data.

The picture shows an example were I'm trying to eliminate data < 1.35(red x-axis). There seems to be loads of other distributions but the information is overwhelming. Would be helpful if someone can point in the right direction.

Thanks.

[Somto Dibiaezue]

Might be worth pointing out the threshold changes with each iteration

[jonas]

Perhaps I'm misunderstanding the issue. Why don't you just find the 20th percentile and remove all values below that?

th=prctile(xdata,20)
xdata(xdata<th)=[];
ydata(xdata<th)=[];

[Somto Dibiaezue]

Thanks a lot