python - Duplicate strings in nested lists -

set - have list nested lists (call relay sets) looks this:

relay_sets = [     [[80100, 'a']],      [[75000, 'b']],      [[64555, 'c']],      [[55000, 'd'], [44000, 'e']],      [[39000, 'f'], [2000, 'g']],      [[2000, 'h'], [999, 'i'], [999, 'j']],      [[999, 'k'], [999, 'l'], [343, 'm']] ] 

a user put in 1 of relay sets (with uniform probability) , again uses 1 of nested lists in sublist equal probability. example user has 1/7 chance of being put in [[80100, 'a']] guaranteed use [80100, 'a'], whereas user has 1/7 chance of being put in [[55000, 'd'], [44000, 'e']] , has 1/2 chance of using [55000, 'd'].

i wrote script finds probability of using relay:

def prob_using_relay(relay, relaysets):   prob1 = 1 / float(len(relaysets))   index, item in enumerate(relaysets):     in item:       if relay in i:         num_relays_in_set = len(relaysets[index])         prob2 = 1 / float(num_relays_in_set)   total_prob = float(prob1 * prob2)   return total_prob  print prob_using_relay(80100, relay_sets) # gives correct answer of 0.142857142857 

my problem don't know how account duplicates. if wanted know chance of using relay 999 bandwidth answer should (2/3 * 1/7 + 2/3 * 1/7) script gives 1/21. or if wanted know chance of using relay 2000 bandwidth answer should (1/2 * 1/7 + 1/3 * 1/7) instead of (1/2 * 1/7).

could altering function account duplicates?

there lots of problems code. doing useless casts floats. being can't handle probabilities more 2 levels deep. here's recursive approach can handle this:

relay_sets = [     [[80100, 'a']],     [[75000, 'b']],     [[64555, 'c']],     [[55000, 'd'], [44000, 'e']],     [[39000, 'f'], [2000, 'g']],     [[2000, 'h'], [999, 'i'], [999, 'j']],     [[999, 'k'], [999, 'l'], [343, 'm']] ]  def prob_using_relay(relay, relaysets):     if isinstance(relaysets[0], int):         return int(relaysets[0] == relay)     return sum(prob_using_relay(relay, relayset) relayset in relaysets) / len(relaysets)  print(prob_using_relay(80100, relay_sets))