Removing Highest and Lowest Measurements to Average

I have a large set of volume measurments I need to sort through. I have about 150 samples and for each sample the instrument measures the volume 10 times and reports all 10. Our procedure is to remove the highest and lowest values and then average the remaining eight. As it stands I've been doing it manually, but with such a large dataset I would prefer to automate it. I'm not even really sure where to start for the matlab code. Selecting out the mins and maxes to make a new list works individually but I don't know how to scale that to 150 samples.

Thanks!

[dpb]

Well, how do you have the data -- individual files, an array 10 measurements by row the by 150 observations, ... ???

It's a trivial MATLAB array operation once you have the data in hand; be we don't know how that is available. Certainly there's no point in doing such tedious things manually.

But, for example, if you did have such an array in which each column was an observation and each row the 10 measurments, then

SND=sort(SND);                  % sort all columns
MEANSND=mean(SND(2:end-1,:));   % compute mean of middle 8 ignoring min/max

Of course, it would then be trivial to also just save the middle eight; you can even save the original order vector and rearrange the remaining elements in their original order if that is significant.

The hardest part undoubtedly is just getting the data into MATLAB to begin with, and that's likely not that difficult, either! :)

[Sophie Leiter]

This was helpful, thank you!

[Image Analyst]

Sophie, did you even see my Answer below? I also showed you how to process a sequence of files like you asked. It also shows you how to remove the min(s) and max(es) from the list. The Comment above will not remove them all if there is more than one min or max.

[Sophie Leiter]

Yes, thank you! I ended up doing a combination of both your answers and it worked well. Thanks for helping me!

Answer 1

[Image Analyst]

To process a sequence of files, see code samples in the FAQ:

https://matlab.fandom.com/wiki/FAQ#How_can_I_process_a_sequence_of_files.3F

Once you've read in the data, you can get the min and max and remove them from the sum:

minValue = min(data(:));
maxValue = max(data(:));
% Min and max may occur more than once.  Find out how many times they occur.
numMins = sum(data(:) == minValue)
numMaxs = sum(data(:) == maxValue)
theSum = sum(data(:) - numMins * minValue - numMaxs * maxValue)
theMean = theSum / (numel(data) - numMins - numMaxs)