>>> lis=[['a','b','c'],['d','e','f']]
>>> any('d' in x for x in lis)
True
generator expression using any
$ python -m timeit -s "lis=[['a','b','c'],['d','e','f'],[1,2,3],[4,5,6],[7,8,9],[10,11,12],[13,14,15],[16,17,18]]" "any('d' in x for x in lis)"
1000000 loops, best of 3: 1.32 usec per loop
generator expression
$ python -m timeit -s "lis=[['a','b','c'],['d','e','f'],[1,2,3],[4,5,6],[7,8,9],[10,11,12],[13,14,15],[16,17,18]]" "'d' in (y for x in lis for y in x)"
100000 loops, best of 3: 1.56 usec per loop
list comprehension
$ python -m timeit -s "lis=[['a','b','c'],['d','e','f'],[1,2,3],[4,5,6],[7,8,9],[10,11,12],[13,14,15],[16,17,18]]" "'d' in [y for x in lis for y in x]"
100000 loops, best of 3: 3.23 usec per loop
How about if the item is near the end, or not present at all? any
is faster than the list comprehension
$ python -m timeit -s "lis=[['a','b','c'],['d','e','f'],[1,2,3],[4,5,6],[7,8,9],[10,11,12],[13,14,15],[16,17,18]]"
"'NOT THERE' in [y for x in lis for y in x]"
100000 loops, best of 3: 4.4 usec per loop
$ python -m timeit -s "lis=[['a','b','c'],['d','e','f'],[1,2,3],[4,5,6],[7,8,9],[10,11,12],[13,14,15],[16,17,18]]"
"any('NOT THERE' in x for x in lis)"
100000 loops, best of 3: 3.06 usec per loop
Perhaps if the list is 1000 times longer? any
is still faster
$ python -m timeit -s "lis=1000*[['a','b','c'],['d','e','f'],[1,2,3],[4,5,6],[7,8,9],[10,11,12],[13,14,15],[16,17,18]]"
"'NOT THERE' in [y for x in lis for y in x]"
100 loops, best of 3: 3.74 msec per loop
$ python -m timeit -s "lis=1000*[['a','b','c'],['d','e','f'],[1,2,3],[4,5,6],[7,8,9],[10,11,12],[13,14,15],[16,17,18]]"
"any('NOT THERE' in x for x in lis)"
100 loops, best of 3: 2.48 msec per loop
We know that generators take a while to set up, so the best chance for the LC to win is a very short list
$ python -m timeit -s "lis=[['a','b','c']]"
"any('c' in x for x in lis)"
1000000 loops, best of 3: 1.12 usec per loop
$ python -m timeit -s "lis=[['a','b','c']]"
"'c' in [y for x in lis for y in x]"
1000000 loops, best of 3: 0.611 usec per loop
And any
uses less memory too