How can I end a for loop when a specific value is reached and create a new array with data that meets that condition?

I'm trying to simulate the loss of the highest values in a dataset. My end goal is to create a new array that matches the mean of observed data -- i.e. I need to calculate means for modeled values in cells 1-2, 1-3, 1-4... 1-n until I find the range of modeled values with a mean that matches my observed data. Right now, I'm trying to work with a for loop that will terminate once a mean threshold is met -- when that threshold is met, I would like to create a new array (rh2 in my code) that incorporates the cells used to calculate the modeled mean that best matches my observed mean. My code seems giving me output of a modeled mean that matches my observed mean, but not the output of the array that I need. Thanks!

meanr2 = mean(ruber2);
ko = fitdist(ruber1,'kernel');
ruber = ko.random(500,1);
sort_ruber = sort(ruber);
for k=1:500
    mur = mean(sort_ruber(1:k));
    if mur <= meanr2
        rh2 = ruber(1:k);
    elseif mur > meanr2
        return
    end
end

Answer 1

[Brigitta Rongstad]

actually your answer helped me figure it out -- thank you! I need to output the sort_ruber rather than the ruber.

Answer 2

[KL]

If I understand correctly, you want to compare meanr2 with the mean calculated from ruber but you're actually comparing it with the mean of sort_ruber which is not the same and then you make a copy from ruber.

when you sort, the values in the array are rearranged.

>> ruber = rand(10,1)

ruber =

      0.7803
      0.3897
      0.2417
      0.4039
      0.0965
      0.1320
      0.9421
      0.9561
      0.5752
      0.0598

>> sort_ruber = sort(ruber)

sort_ruber =

      0.0598
      0.0965
      0.1320
      0.2417
      0.3897
      0.4039
      0.5752
      0.7803
      0.9421
      0.9561

so the mean you calculate from 1:k is not the same for both vectors. Change

mur = mean(sort_ruber(1:k));

to

mur = mean(ruber(1:k));

then it should give you what you want.

Answer 3

[Brigitta Rongstad]

I need to calculate the mean iteratively by removing the highest value one at a time (or by adding a higher value one at a time) -- I sorted because this was the way I could think of trying to calculate a mean for values lowest-second lowest, lowest-third lowest... on until I'm calculating the mean for the lowest-highest values. Once I find the range of values (i.e. lowest to highest value range that gives me the mean I'm looking for), I need to output that value range.