年度別法人番号付与日.py

import pandas as pd

csv_file = '00_zenkoku_all_20240830.csv'
df = pd.read_csv(csv_file, header=None, encoding='utf-8', low_memory=False)

df.columns = [
    'sequence_number', 'corporate_number', 'process', 'correct', 'update_date',
    'change_date', 'name', 'name_image_id', 'kind', 'prefecture_name', 'city_name',
    'street_number', 'prefecture_code', 'city_code', 'postcode', 'address_outside',
    'address_outside_image_id', 'close_date', 'close_cause', 'successor_corporate_number', 
    'change_cause', 'extra_column', 'assignment_date', 'latest', 'name_en', 'prefecture_name_en', 'city_name_en', 
    'address_outside_en', 'furigana', 'hihyoji'
]

new_establishments = df[df['process'] == 1]
new_establishments['assignment_date'] = pd.to_datetime(new_establishments['assignment_date'], errors='coerce')

def calculate_fiscal_year(date):
    if pd.isnull(date):
        return None
    if date.month >= 4:
        return date.year
    else:
        return date.year - 1

new_establishments['fiscal_year'] = new_establishments['assignment_date'].apply(calculate_fiscal_year)
new_establishments = new_establishments.dropna(subset=['fiscal_year'])
result = new_establishments.groupby(['city_name', 'fiscal_year']).size().reset_index(name='count')

print(result)

result.to_csv('new_establishments_by_city_fiscal_year.csv', index=False, encoding='utf-8')