7

I want to ask you about the notation in probability.

I know that

P(A | B) = the conditional probability that event A occurs given that event B has occurred already

But I cannot find what A,B or in my case P(A|B,C). I suggest it means "the conditional probability that event A occurs given that B and C BOTH occurred already"

I don't know what the comma means.

Can you help me ?

Stephane Rouberol
  • 4,286
  • 19
  • 18
user1328370
  • 401
  • 1
  • 3
  • 6

2 Answers2

11

You are basically correct.

P(A| B) is the probability of A given B. P(A| B, C) is the probability of A given (B and C).

You could just as easily write it as P(A| B ∧ C) but it is notational convention to use a comma. Think of everything after the vertical bar as a list of the given things, separated by commas.

(And note that the vertical bar is a very high precedence operator, so to speak.)

pglpm
  • 516
  • 4
  • 14
Novak
  • 4,687
  • 2
  • 26
  • 64
  • 2
    And is there a notation for "the probability of B or C, givven that A happened": P(A|[B or C]), not P(A,[B and C]) - just curious – user1328370 Apr 23 '14 at 14:29
  • "The probability of *B* or *C*, given *A*" is denoted P(*B* ∨ *C* | *A*). – pglpm May 08 '18 at 08:32
  • 4
    Isn't it lower order of precedence, since you apply the ANDs for the whole list before you apply the given-bar? – dekuShrub Sep 02 '20 at 06:55
  • @dekuShrub you are correct, the author of the answer uses the words "high precedence" in the wrong way here. I think what he actually means is "giving the expression structure on a higher level than the commas", which basically means it's evaluated later than the commas, hence it has lower precedence, not higher precedence. – Daniel S. Jul 28 '23 at 08:20
  • @Novak can you also give an example with many more variables and tell what it means, e.g. P(A, B, C, D, E | F, G, H, I, J, K)? I think it gets clearer then. – Daniel S. Jul 28 '23 at 08:23
1

This is according to Bayes rule

P(C|A,B) = P(A,B|C).P(C) / P(A,B)

amitchhajer
  • 12,492
  • 6
  • 40
  • 53