detailed_sorting_explanation.py

counter = {}
for word in words:
    counter[word] = counter.get(word, 0) + 1      # we basically count the words in the dict e.g. {'mickey': 1, 'alex': 2, 'anna': 2}
    freq = list(counter.keys())                   # we set up a list e.g. freq = ['mickey', 'alex', 'anna']
    freq.sort(key = lambda x : (-counter[x], x))  # the intention here is to sort by frequency in descending oder 2,1,0...

    # .sort (https://docs.python.org/3.3/library/stdtypes.html#list.sort)
    # Sort the freq list in place
    # takes a key-word argument called 'key' to be used for a comparison function
    # lambda (https://docs.python.org/3/reference/expressions.html#lambda)
    # is an anonymous function (meaning it doesn't have a name)
    # you could have done def compare(key): return (-counter[x], x) but you can see it's more verbose
    # -counter[x] -> this basically takes a key like 'mickey' and grabs the associated value 1
    # and negates it so it becomes -1, we do this so we can have descending order (-2, -1, 0)
    # as by default the sorting is in ascending order (0, 1, 2); this can be achieved by passid the 'reversed' keyword arg
    # The function returns a tuple (-counter[x], x), this is so that
    # if we have the same frequency (-2, 'mickey'), (-2, 'anna') we can sort on the second element
    # Tuples are sorted lexicographically, starting with the first item then the second etc.
    # the resulting freq list will be sorted by frequency of occurrence followed by lexicographically
    # so ['alex', 'anna', 'mickey']