import random
import collections
 
PERCENTAGE = 0.2
 
names_all = ["albert", "bruce", "anna", "bella", "cindy", 
	"colbert", "anakin", "al gore", "carla", "cole"] * 50 # men + women, men.extend(women), etc
 
names_subset = random.sample(names_all, int(PERCENTAGE * len(names_all)))
 
names_dict = collections.defaultdict(list)
for name in names_subset:
	names_dict[name[0].lower()].append(name)
 
for letter in names_dict:
	names = names_dict[letter]
	names_dict[letter] = random.sample(names, int(PERCENTAGE * len(names)))
 
print(names_dict)

aW1wb3J0IHJhbmRvbQppbXBvcnQgY29sbGVjdGlvbnMKClBFUkNFTlRBR0UgPSAwLjIKCm5hbWVzX2FsbCA9IFsiYWxiZXJ0IiwgImJydWNlIiwgImFubmEiLCAiYmVsbGEiLCAiY2luZHkiLCAKCSJjb2xiZXJ0IiwgImFuYWtpbiIsICJhbCBnb3JlIiwgImNhcmxhIiwgImNvbGUiXSAqIDUwICMgbWVuICsgd29tZW4sIG1lbi5leHRlbmQod29tZW4pLCBldGMKCm5hbWVzX3N1YnNldCA9IHJhbmRvbS5zYW1wbGUobmFtZXNfYWxsLCBpbnQoUEVSQ0VOVEFHRSAqIGxlbihuYW1lc19hbGwpKSkKCm5hbWVzX2RpY3QgPSBjb2xsZWN0aW9ucy5kZWZhdWx0ZGljdChsaXN0KQpmb3IgbmFtZSBpbiBuYW1lc19zdWJzZXQ6CgluYW1lc19kaWN0W25hbWVbMF0ubG93ZXIoKV0uYXBwZW5kKG5hbWUpCgpmb3IgbGV0dGVyIGluIG5hbWVzX2RpY3Q6CgluYW1lcyA9IG5hbWVzX2RpY3RbbGV0dGVyXQoJbmFtZXNfZGljdFtsZXR0ZXJdID0gcmFuZG9tLnNhbXBsZShuYW1lcywgaW50KFBFUkNFTlRBR0UgKiBsZW4obmFtZXMpKSkKCnByaW50KG5hbWVzX2RpY3Qp