load_weather.py

import numpy as np
import pandas as pd
import time

base_url = 'https://www.wunderground.com/history/airport/KBOS/{}/{}/{}/DailyHistory.html?&reqdb.zip=&reqdb.magic=&reqdb.wmo=&format=1'

d_range = pd.to_datetime(np.arange('2015-12-29', '2016-06-15', dtype='datetime64[D]'))

print(d_range)

result = pd.DataFrame()
for date in d_range:
    url = base_url.format(date.year, date.month, date.day)
    d_csv = pd.read_csv(url, header=0, dtype='str')
    result = result.append(d_csv, ignore_index=True)
    print(base_url.format(date.year, date.month, date.day))
    print(d_csv.shape[0])
    time.sleep(2)


fname = 'data/weather_' + str(d_range.min().strftime('%Y-%m-%d')) + '_' + str(d_range.max().strftime('%Y-%m-%d')) + '.csv'
result.to_csv(fname, index=False)
print(result)