igm_validation.py

# Correct response types
df.loc[df['response'].str.contains(r'did not', case=False) | df['response'].str.contains(r'---'), 'response'] = np.nan

# Convert empty string comments into null types
df['comment'] = df['comment'].replace(r'^$', np.nan, regex=True)

# Assign sex variable to economists
sex = pd.read_csv(os.path.join(os.path.dirname(__file__), 'economist_sex_mapping.csv'), index_col='economist_name')
df['sex'] = df['economist_name'].map(sex['sex'])

# Assign response categories to numerical values    
certainty_mapping = {
    'Strongly Disagree': -2, 
    'Disagree': -1,
    'Uncertain': 0,
    'No opinion': 0,
    'Agree': 1,
    'Strongly Agree': 2,
}
df = df.assign(response_int = lambda x: x['response'].map(certainty_mapping))