def collect_google_scores(terms, start, end):
logging.getLogger('googleapiclient.discovery_cache').setLevel(logging.ERROR)
logger.info('Querying %d terms between %s and %s' % (len(terms), start, end))
logger.debug(', '.join(t.term for t in terms))
service = build(
'trends',
'v1beta',
developerKey=os.environ["GOOGLE_API_KEY"],
discoveryServiceUrl='https://www.googleapis.com/discovery/v1/apis/trends/v1beta/rest',
cache_discovery=False
)
graph = service.getTimelinesForHealth(
terms=[t.term for t in terms],
geoRestriction_region='GB-ENG',
time_startDate=start.strftime('%Y-%m-%d'),
time_endDate=end.strftime("%Y-%m-%d"),
timelineResolution='day')
try:
response = graph.execute()
except HttpError as e:
logger.exception(e)
raise e
for line in response['lines']:
term = next(t for t in terms if t.term == line['term'])
for point in line['points']:
day = datetime.strptime(point['date'], "%b %d %Y").date()
gs = get_google_score(term, day)
gs.value = float(point['value'])
yield gs
def find_matching_index(headers, possible, required=False):
for i, h in enumerate(headers):
if h in possible:
return i
if required:
e = FluDetectorError('No %s header in CSV file' % ','.join(possible))
logger.exception(e)
raise e
def find_region_index(headers):
indexes = {}
for code, name in REGIONS.iteritems():
index = find_matching_index(headers, [code, name])
if index:
indexes[code] = index
if not indexes:
e = FluDetectorError('No region headers found')
logger.exception(e)
raise e
return indexes
def calculate_score(model, day, engine_runner):
averages = list(calculate_moving_averages(model, day))
if not averages:
return
ms = get_model_score(model, day)
try:
if engine_runner.conf is CalculatorType.LEGACY:
ms.value = send_to_matlab(model, averages)
else:
ms.value = engine_runner.calculateModelScore(model, averages)
except ErrorReturnCode as e:
logger.exception(e)
raise e
return ms