Распределение элементов по группам без пересечений.

gistfile1.py

#!/usr/bin/env python

from __future__ import print_function
import collections

GROUPS_COUNT = 7

cleaned_words = set()

#
# Prepare input
#
for word in open('/tmp/doc.txt', 'rb').read().decode('utf-8').strip().split():
    if len(word) < 5:
        continue
    if not word[0].isalpha():
        continue
    cleaned_words.add(word.lower())

words_grouped_by_first_letter = collections.defaultdict(list)

for word in cleaned_words:
    words_grouped_by_first_letter[word[0]].append(word)


wordscount_grouped_by_first_letter = collections.defaultdict(list)
for letter, words in words_grouped_by_first_letter.items():
    wordscount_grouped_by_first_letter[letter] = len(words)

result = []

sorted_letters = sorted(wordscount_grouped_by_first_letter.keys())
result = [[letter] for letter in sorted_letters]

print("==> Source ======> ")
for letter in sorted_letters:
    print(letter, wordscount_grouped_by_first_letter[letter])

#
# Process
#
print("==> Merging =====> ")
while len(result) > GROUPS_COUNT:
    min_val = float('inf')
    min_idx = None
    for index in range(len(result)-1):
        cur_el_weight = sum(
            wordscount_grouped_by_first_letter[key]
            for key in result[index]
        )
        next_el_weight = sum(
            wordscount_grouped_by_first_letter[key]
            for key in result[index+1]
        )
        if (cur_el_weight + next_el_weight) < min_val:
            min_idx = index
            min_val = cur_el_weight + next_el_weight

    if min_idx is not None:
        result[min_idx] += result[min_idx+1]
        del result[min_idx+1]

#
# Print result
#
print("==> Result ======> ")

for keys in result:
    print(
        keys[0], "-", keys[-1],
        sum([wordscount_grouped_by_first_letter[key] for key in keys]),
    )