Donald Trump Gallup approval rating weighted by party identification

data.csv

make_csv.py

'''
Script to parse Trump approval (sourced from Gallup website, 
table copy+paste from browser), weighted by party ID, and output as CSV
'''

import csv
import datetime
import string

#Republicans Independents    Democrats
PARTY_ID = '''
2017 Jul 5-9    25  45  28
2017 Jun 7-11   26  42  30
2017 May 3-7    29  40  28
2017 Apr 5-9    25  44  28
2017 Mar 1-5    26  42  30
2017 Feb 1-5    31  37  31
2017 Jan 4-8    28  44  25
'''

#Republicans + Republican leaners    Democrats + Democratic leaners
INDIE_LEANERS = '''
2017 Jul 5-9    40  48
2017 Jun 7-11   43  49
2017 May 3-7    45  44
2017 Apr 5-9    41  48
2017 Mar 9-29   38  47
2017 Mar 1-5    41  49
2017 Feb 1-5    43  48
2017 Jan 4-8    44  43
'''

#Republicans Independents    Democrats
TRUMP_APPROVAL = '''
2017 Jul 24-30  82  32  7
2017 Jul 17-23  86  31  8
2017 Jul 10-16  87  33  8
2017 Jul 3-9    85  35  8
2017 Jun 26-Jul 2   85  36  8
2017 Jun 19-25  85  34  6
2017 Jun 12-18  84  32  6
2017 Jun 5-11   83  31  8
2017 May 29-Jun 4   82  34  7
2017 May 22-28  87  37  8
2017 May 15-21  84  31  7
2017 May 8-14   84  35  8
2017 May 1-7    84  40  9
2017 Apr 24-30  87  36  12
2017 Apr 17-23  86  36  9
2017 Apr 10-16  87  36  10
2017 Apr 3-9    87  34  6
2017 Mar 27-Apr 2   81  36  6
2017 Mar 20-26  84  33  8
2017 Mar 13-19  86  35  10
2017 Mar 6-12   88  36  9
2017 Feb 27-Mar 5   88  39  10
2017 Feb 20-26  88  38  10
2017 Feb 13-19  86  37  7
2017 Feb 6-12   87  35  11
2017 Jan 30-Feb 5   86  41  8
2017 Jan 20-29  89  42  13
'''

from_date_fmt = '%Y %b %d'
to_date_fmt_no_month = '%d'
to_date_fmt_with_month = '%b %d'

def get_range_from_line(line):
    date_range_str = ' '.join(line.split()[0:3])
    from_date_str = date_range_str.split('-')[0]
    from_date = datetime.datetime.strptime(from_date_str, from_date_fmt).date()
    after_dash = line.split('-')[1]
    if after_dash[0] in string.letters:
        to_date_str = ' '.join(after_dash.split()[0:2])
        to_date_month_day = datetime.datetime.strptime(
                to_date_str, to_date_fmt_with_month
        )
        to_date = datetime.date(
                from_date.year, to_date_month_day.month, to_date_month_day.day
        )
    else:
        to_date_str = after_dash.split()[0]
        to_date_day = datetime.datetime.strptime(to_date_str, to_date_fmt_no_month)
        to_date = datetime.date(
                from_date.year, from_date.month, to_date_day.day
        )
    return (from_date, to_date)

def parse_id_approval(which):
    output = []
    for line in which.split('\n'):
        if line == '':
            continue
        date_range = get_range_from_line(line)
        rep, ind, dem = line.split('  ')[-3:]
        output.append({
                'from': date_range[0], 
                'to': date_range[1], 
                'rep': int(rep)/100.0,
                'ind': int(ind)/100.0, 
                'dem': int(dem)/100.0,
        })
    return tuple(output)

def make_weighted_approvals(id_ranges, approval_ranges):
    output = []
    for poll in approval_ranges:
        which_id_range = None
        for i, id_range in enumerate(id_ranges):
            if id_range['from'] < poll['to']:
                index = i-1 if i > 0 else i
                which_id_range = id_ranges[i]
                break
        output.append({
                'from': poll['from'],
                'to': poll['to'],
                'weighted_rep': poll['rep'] * which_id_range['rep'],
                'weighted_ind': poll['ind'] * which_id_range['ind'],
                'weighted_dem': poll['dem'] * which_id_range['dem'],
                'unweighted_rep': poll['rep'],
                'unweighted_ind': poll['ind'],
                'unweighted_dem': poll['dem'],
        })
    return output

def write_csv():
    id_ranges = parse_id_approval(PARTY_ID)
    approval_ranges = parse_id_approval(TRUMP_APPROVAL)
    weighted_approvals = make_weighted_approvals(id_ranges, approval_ranges)

    with open('weighted_trump_approval.csv', 'w') as fout:
        writer = csv.writer(fout)
        keys = ('from', 'to', 'weighted_rep', 'weighted_ind', 'weighted_dem', 
                'unweighted_rep', 'unweighted_ind', 'unweighted_dem'
        )
        writer.writerow(keys)
        for approval in weighted_approvals:
            writer.writerow([approval[k] for k in keys])

if __name__ == '__main__':
    write_csv()