Single-malt scotch

0_preprocess.py

import csv
import json

rd = csv.reader(open('scotch.csv'))
header = rd.next()
colors = header[1:15]
data = []
schema = {
    'color': { 'type': 'categorical' },
    'AGE': { 'type': 'count' },
    'SCORE': { 'type': 'count' },
    'REGION': { 'type': 'categorical' },
    'DISTRICT': { 'type': 'categorical' }
}
for line in rd:
    row = {}
    row['_id'] = line[0].replace('-','').replace(' ','_')
    print row['_id']
    for i, x in enumerate(line[1:15]):
        if int(x) == 1:
            row['color'] = colors[i]
    for i, x in enumerate(line):
        if header[i].startswith('nose_') or header[i].startswith('body_') or header[i].startswith('palette_') or header[i].startswith('finish_'):
            row[header[i]] = bool(int(x))
            schema[header[i]] = {'type': 'boolean'}
        elif header[i] == 'AGE':
            if x != '':
                row['AGE'] = int(x)
        elif header[i] == 'SCORE':
            row['SCORE'] = int(x)
        elif header[i] == 'REGION':
            row['REGION'] = x
        elif header[i] == 'DISTRICT':
            row['DISTRICT'] = x
    data.append(row)

open('scotch.json','wb').write(json.dumps(data, indent=2))
open('schema.json','wb').write(json.dumps(schema, indent=2))
    

1_run.py

import veritable
import veritable.utils
import json

data = json.loads(open('scotch.json').read())
schema = json.loads(open('schema.json').read())
veritable.utils.clean_data(data, schema)

api = veritable.connect()
table = api.create_table('scotch', force=True)
table.batch_upload_rows(data)
analysis = table.create_analysis(analysis_id='scotch', schema=schema)

2_scotch.csv