def _write_language_file(self, out_file):
coords = Coordinates()
langs = defaultdict(lambda: {'count': 0, 'group': None})
for year, entry_list in self.groups:
if year >= START_YEAR and year <= END_YEAR:
entry_list = list(entry_list)
for entry in entry_list:
langs[entry.language]['count'] += 1
langs[entry.language]['group'] = entry.language_group_initial()
for language in langs.keys():
# Number of possible points (between 3 and 30, depending on
# the frequency of the language)
num_points = int(langs[language]['count'] / 5)
num_points = max(4, min(num_points, 30))
# Select a bunch of random points within the language's geo region
langs[language]['coords'] = [coords.randomize(
language, decimalPlaces=2) for i in range(num_points)]
langs2 = []
for language, vals in langs.items():
langs2.append({'l': language,
'g': vals['group'],
'c': vals['coords']})
with open(out_file, 'w') as filehandle:
json.dump(langs2, filehandle)
language_index = {row['l']: i for i, row in enumerate(langs2)}
return language_index
