Gendercounter

A program for counting the occurrence of male and female Swedish names in a text file.

Try the live web version: genuskollen.se

Functionality

A series of functions for counting gendered Swedish names in text. Import the gendercounter.py script along with the tsv files containing names and frequencies:

female250.tsv
male250.tsv

The first two dictionaries list female and male names respectively, along with the frequency of occurrence according to Statistics Sweden (2015). Also, the following pronouns can be counted.

hen/henom/hens
hon/henne/hennes
han/honom/hans

Usage

import gendercounter # Just import as usual.

example = """Karl och Lisa promenerar på gatan och
            tycker om Ove och Stina och Bertil.
            Hon, han, hen, henom. Hon hens"""

Usage example 1: initialize a text object from a string

text = gendercounter.from_string(example)

text.genderfrequency()

{'Men': 3, 'Women': 2}

text.pronounfrequency()

{'han': 1, 'hans': 0, 'hen': 1, 'henne': 0, 'hennes': 0, 'henom': 1, 'hens': 1, 'hon': 2, 'honom': 0}

# Returns names and the number of people with that name in Sweden.
text.names()

{'Men': {'Bertil': '68261', 'Karl': '209908', 'Ove': '33731'}, 'Women': {'Lisa': '31611', 'Stina': '19071'}}

Usage 2: From a text file

textfile = gendercounter.from_textfile("testtext.txt")

textfile.genderfrequency()

{'Men': 15, 'Women': 7}

Usage 3: Do both as a single line of code

gendercounter.from_textfile("testtext2.txt").genderfrequency()

{'Men': 43, 'Women': 29}

Usage 4: Iterate over multiple files, example

from os import listdir

for file in listdir('.'):
    if file.endswith('txt'):
        print(file)
        for k, v in gendercounter.from_textfile(file).genderfrequency().items():
            print(k, v)
        for k, v in gendercounter.from_textfile(file).pronounfrequency().items():
            print(k, v)
        print("-" * 10)

testtext2.txt Men 43 Women 29 han 1 henom 0 hennes 0 henne 0 hon 0 hen 0 honom 0 hens 0 hans 0

testtext.txt Men 15 Women 7 han 1 henom 2 hennes 2 henne 2 hon 2 hen 2 honom 0 hens 2 hans 0

Concurrency

If you work with large datasets, check out the concurrency.ipynb notebook for an example of concurrent computing.

Sources of error

Some names are also frequent Swedish words, for example "De".
Uncommon names (less than 250 occurrences) were excluded.
Some names are gender neutral ("Charlie", "Mario", "Alex" etc.)
With the pronoun counter, the words "Han" and "Hans" can also be Swedish male names.
??? (Please let me know if you find other sources of error)

Removed names

"De"
"Del"

Similar programs

SexMachine (English language, written in Python)

christopherkullenberg/gendercounter