Calculate min/max/mean value every 100 points

Hey! I have an array with 17 999 elements and for every 100 points I want to calculate min, max and mean values. First I thought about using reshape and convert the array into a table with 100 rows and just calculate what I need for every column but it doesn't work with the number of elelmts that I have. I am wondering if there is any other way?

[Image Analyst]

Does your window of 100 elements slide along one element at a time (see movmean, etc), or does it move in "jumps" of 100 elements? What are your plans for the last 99 points? Why is your vector not a multiple of 100?

Answer 1

[Rik]

Reshape will work, but you have to pad with NaN values and set the 'omitnan' flag:

data=rand(1,17999);
data(end+(1:mod(-end,100)))=NaN;
data=reshape(data,100,[]);
mean(data,'omitnan')

ans = 1×180

0.5982 0.4930 0.4950 0.5386 0.4964 0.5032 0.5025 0.4941 0.5585 0.5125 0.4598 0.5637 0.4945 0.5666 0.5186 0.5374 0.4860 0.4730 0.5227 0.4853 0.5401 0.4786 0.5643 0.4976 0.4568 0.4634 0.4806 0.5300 0.4638 0.4862

Answer 2

[David Hill]

Just remove the last 99 elements. If you need the mean, max, min for the last 99 elements, calculate them separately.

r=reshape(yourMatrix(1:17900),100,[]);
Min=min(r);
Max=max(r);
Mean=mean(r);

Answer 3

[William Rose]

@Liliya Vasilevich,

N=17999;
M=1000;
P=floor(N/M);
x=rand(1,N);
xavg=zeros(1,P); %allocate array for average values
for i=1:P
    xavg(i)=mean(x((i-1)*M+1:i*M));
end
plot(1:P,xavg,'-r*')

Try it. Good luck.

[William Rose]

@Liliya Vasilevich, Now I see that you requested every 100 points, and I wrote a script for 1000 points. Change M from 1000 to 100 to get the average every 100 points.