Finding the Probability of a Sequence of Numbers

Hi everyone,

I have a vector with data varying from 1 to 120.

A= [10 13 10 1 111 102 10 13 7 10 112 100 14 116 102 10 14 120].

I would like to calculate, the probability' sequence of each number (e.g., probability of the number 1 be followed by the number 111; probability of the number 14 be followed by the number 116, etc ...)?

How can I do that?

Thanks.

[Torsten]

Depends on how A was created.

[Adam]

Sounds like you would need to build a 120x120 matrix of occurences of each number followed by each other number (something I guess I would use a for loop to do, at least initially) and then work out probabilities from the sum of each row in the table. It's a very small sample to do any kind of probability analysis on though. I assume this is just an example and your real data is bigger?

[Aragorn23]

You're right, it's just a small sample.

Answer 1

[Walter Roberson]

A= [10    13    10    1    111    102    10    13    7    10    112    100    14    116    102    10    14    120];
Ac = A.';
counts = accumarray([Ac(1:end-1), Ac(2:end)], 1, [120 120], [], [], true);

probs = counts ./ max(1, sum(counts,2));

[c, r, s] = find(probs.');
[r, c, s]

[Aragorn23]

Thank you Walter Roberson.

Answer 2

[Bruno Luong]

N = histcounts2(A(1:end-1),A(2:end),1:121,1:121);
p = N / sum(N(:))