I want to count (tally) the number of occurrences of any generated integer over a 156 for-loop.

I want to run a iteration 156 times via a for-loop. It will randomly generate an array of 5 numbers between 1-100. I want to know how many time any number within the range occured over the 156 iteration for loop. How would I be able to accomplish this?

Bonus question: Is there a way to code how many integers appear together in any given array?

Perhaps there is an easier way to code this type of iteration?

[Torsten]

And what have you done so far ?

[Jeffrey Lewis]

Sorry, I forgot to add what I created. It's not much at all but this is where I am with the for-loop

Ultimately, I would like the code to look at each array output and count multiple occurrences of any integer within the range set and output the occurrence in a tallied list, ideally

clc
clear all

for K = 1:156
    n = randi([1 100],1,5)
end

[Torsten]

rng('default')
edges = 1:100;
data = randi(100,156,5);
count = arrayfun(@(i)nnz(data(:)==edges(i)),edges);
count

count = 1×100

4 7 4 10 9 9 6 5 10 10 12 8 8 9 9 6 9 11 11 12 7 3 8 8 7 9 12 5 7 9

sum(count)

ans = 780

156*5

ans = 780

[David Hill]

data = randi(100,1,156*5);
h=histc(data,1:100)

h = 1×100

8 7 4 17 9 12 9 5 5 9 6 11 8 9 7 8 8 6 8 10 12 7 6 7 5 4 8 8 6 8

Answer 1

[Steven Lord]

Are all these values integer values? If so consider using histcounts with the 'integers' BinMethod.

values = randi(100, [200, 5]);
[counts, edges] = histcounts(values, BinMethod='integers', BinLimits = [1 100]);
spotCheck = [counts(42), nnz(values == 42)]

spotCheck = 1×2

10 10

We can also display a histogram and plot the data from histcounts to check.

histogram(values, BinMethod='integers', BinLimits = [1 100]);
hold on
bincenters = (edges(1:end-1)+edges(2:end))./2;
plot(bincenters, counts, 'ro')
plot(42, counts(42), 'k+')

It's a little hard to see the black + in the small picture on Answers, but if you ran this code in MATLAB and zoomed in you'd see it more clearly.

Answer 2

[Walter Roberson]

integers appearing "together" is not clear. Since your generation is an ordered vector, does that mean that the integers must be immediately beside each other in sequence? Or does it mean that as long as they show up together in the same vector of 5 that you want it to be counted?

For example, [2 7 7 3 1] -- should that increment the counts for (2,7), (7,7), (7,3), (3,1) ? Or should it increment the counts for (2,7), (7,2), (7,7), (7,3), (3,7), (3,1), (1,3) ? Or (2,7), (7,2), (7,7), (7,3), (3,7), (3,1), (1,3) and another (7,7) as well? Or for (1,2), (1,3), (1,7), (2,3), (2,7), (7,7) ? Or for (1,2), (1,3), (1,7), (2,1), (2,3), (2,7), (3,1), (3,2), (3,7), (7,1), (7,2), (7,3), (7,7) ? Or for (1,2), (1,3), (1,7), (1,7), (2,1), (2,3), (2,7), (2,7), (3,1), (3,2), (3,7), (3,7), (7,1), (7,2), (7,3), (7,7), (7,7) ?

maxval = 100;
counts = zeros(maxval, 1);
paircounts = zeros(maxval, maxval);

for K = 1:156
    n = randi([1 maxval], 5, 1);
    counts = counts + accumarray(n, 1, [maxval 1]);
    paircounts = paircounts + accumarray( [n(1:end-1), n(2:end)], 1, [maxval, maxval]);
end
bar(counts)

imagesc(paircounts); colorbar