Differences between Numpy & List Comprehensions

list_numpy_compare.py

def gen_tickers(num_tickers, num_new):
    """
    This function generates `num_tickers` of random, length 3 tickers and 
    `num_new` random, length 3 tickers
    """
    s  = 'ABCDEFGHIJKLMNOPQRSTUVWXYZ'
    r = numpy.random.randint(1, 27, 3)
    ticks = []
    #create the list of tickers
    for i in numpy.arange(num_tickers):
        ticks.append(
            reduce(lambda x, y: x + y, map(lambda x: s[x], numpy.random.randint(
                0, 26, 3) )) )

    new_ticks = []
    #create the list of tickers to append
    for i in numpy.arange(num_new):
        new_ticks.append(
            reduce(lambda x, y: x + y, map(lambda x: s[x], numpy.random.randint(
                0, 26, 3) )) )

    return ticks, new_ticks

def list_method(ticks, new_ticks):
    """
    Uses list comprehensions to accomplish the goal
    """
    bool_map = map(lambda x: x in ticks, new_ticks)
    return [tick for (tick, bool_val) in zip(new_ticks, bool_map) if not bool_val]

def np_one_method(ticks, new_ticks):
    """
    Uses `numpy`, and simply takes lists as arguments to the 
    `numpy` function
    """
    return numpy.setdiff1d(new_ticks, ticks )

def np_too_method (ticks, new_ticks):
    """
    Uses `numpy`, but converts the lists to `numpy.ndarrays` before
    running the searches
    """
    return numpy.setdiff1d(numpy.array(new_ticks), numpy.array(ticks) )