zip_code = zip_json[u'attributes'][u'Join_Zip_Code']
case_counts.append(case_count)
zip_codes.append(zip_code)
# ===========================================
# Store previous data in dictionary
# ===========================================
us_cases_rel_scripts_path = os.path.abspath(
"../processed_data/cases/US") # Path to data relative to scripts
prev_data_fpath = "%s/oakland-county_cases.csv" % us_cases_rel_scripts_path
prev_data_file = open(prev_data_fpath, 'r')
header = prev_data_file.readline().strip().split(',')
# IMPORTANT: Only append new data if date isn't already there
quoted_date = parse_data_utils.date_string_to_quoted(date)
if quoted_date in header:
prev_data_file.close()
sys.exit(
"This date has already been updated. May need to check if multiple updates were made in the same day."
)
data = {} # Keys are zip codes, values are lists of case counts
for row in prev_data_file:
values = row.strip().split(',')
zip_code = values[0]
case_counts_prev = values[1:]
data[zip_code] = case_counts_prev
def write_data(file_path, date, zips, cases, location=""):
"""Writes or appends the data to a csv. Rows are zip codes and columns are dates.
Args:
file_path: The path to the csv to write/append to.
date: The date we should add data to.
zip: List of zip codes to be updated.
cases: List of cases corresponding to the given list of zip codes and the given date.
location: The location for which the data is being scraped.
"""
# Fetch given csv and store in a dict.
new_data = {} # Keys are zip codes, values are lists of case counts
dates = [""]
# IMPORTANT: Only append new data if date isn't already there
quoted_date = parse_data_utils.date_string_to_quoted(date)
overwrite = False
try:
read_csv = open(file_path, 'r')
dates = read_csv.readline().strip().split(',')
if quoted_date in dates:
overwrite = True
print("%s - WARNING: data has already been updated today... "
"overwriting data for today with recently fetched data." %
location)
# Fill in new_data with old data from the csv.
for row in read_csv:
values = row.strip().split(',')
zip_code = values[0]
old_cases = values[1:]
if overwrite:
old_cases = values[1:-1]
new_data[zip_code] = old_cases
read_csv.close()
except IOError:
print("%s - Creating csv file at %s" % (location, file_path))
num_dates = len(dates) - 1 # Excluding today
if overwrite:
num_dates = num_dates - 1
expected_entries = num_dates + 1
# Fill in new_data with today's data.
for zip_code, case_count in zip(zips, cases):
quoted_zip_code = '"%s"' % zip_code
if quoted_zip_code in new_data:
if len(new_data[quoted_zip_code]) == expected_entries:
new_data[quoted_zip_code][-1] = str(
int(new_data[quoted_zip_code][-1]) + case_count)
else:
new_data[quoted_zip_code].append(str(case_count))
else: # New zip code
new_data[quoted_zip_code] = ['NA'] * (num_dates)
new_data[quoted_zip_code].append(str(case_count))
# Handle missing zip codes
for zip_code in new_data:
if len(new_data[zip_code]) < expected_entries:
new_data[zip_code].append("NA")
# Make sure everything has the same length
# (There might be duplicates!)
for zip_code in new_data:
if len(new_data[zip_code]) != expected_entries:
sys.exit(
"%s - ERROR: Inconsistency in number of data points for zip code %s."
" Data failed to update." % (location, zip_code))
# Overwrite csv
write_csv = open(file_path, 'w+')
if not overwrite:
dates.append(quoted_date)
write_csv.write(','.join(dates) + "\n")
sorted_zip_codes = sorted(new_data.keys())
for zip_code in sorted_zip_codes:
write_csv.write('%s,' % zip_code)
write_csv.write(','.join(new_data[zip_code]) + "\n")
write_csv.close()
