Bug Report

Describe the bug
xtile = int(math.floor((x + EPSILON) * Z2))
ValueError: cannot convert float NaN to integer

To Reproduce
This is the code I am using to collect th close-to images to a specific coordinates:
for index, row in enumerate(df.iterrows()):
coordinates_to_image_id_dict: dict = {}
latitude, longitude = dict(row[1])['LATITUDE'], dict(row[1])['LONGITUDE']
coordinates_to_image_id_dict['lat'] = latitude
coordinates_to_image_id_dict['lng'] = longitude
try:
data= mly.get_image_close_to(
latitude=latitude,
longitude=longitude,
radius=35,
image_type="flat",
min_captured_at="2016-01-01"
).to_dict()

except AttributeError:
print('AttributeError occurred')
data = {}

print(data)
print('latitude and longitude:', latitude, longitude)

Expected behavior
if math.isnan(x):
# handle NaN case here
else:
xtile = int(math.floor((x + EPSILON) * Z2))

Screenshots
latitude and longitude: 51.9326934493218 -8.05201349824569
Requesting GET to https://tiles.mapillary.com/maps/vtp/mly1_public/2/14/7892/5318/?access_token=MLY%???
Response 200 OK received in 78ms
{'type': 'FeatureCollection', 'features': []}

latitude and longitude: 53.2429118794892 -6.58873419758931
Traceback (most recent call last):
File "Data_collection_json_metadata.py", line 39, in
min_captured_at="2016-01-01"
File "/home/peoplexxx/.local/lib/python3.7/site-packages/mapillary/utils/auth.py", line 59, in wrapper
return f(*args, **kwargs)
File "/home/people/xx/.local/lib/python3.7/site-packages/mapillary/interface.py", line 116, in get_image_close_to
kwargs=kwargs,
File "/home/people/xx/.local/lib/python3.7/site-packages/mapillary/controller/image.py", line 110, in get_image_close_to_controller
latitude=latitude,
File "/home/people/xxx/.local/lib/python3.7/site-packages/mapillary/models/api/vector_tiles.py", line 123, in fetch_layer
tile=mercantile.tile(lng=longitude, lat=latitude, zoom=zoom),
File "/home/people/xxxx/.local/lib/python3.7/site-packages/mercantile/init.py", line 426, in tile
xtile = int(math.floor((x + EPSILON) * Z2))
ValueError: cannot convert float NaN to integer

Commencez par mapillary/controller/image.py à get_image_close_to_controller et par mapillary/models/api/vector_tiles.py à fetch_layer, puis reproduisez les coordonnées signalées via get_image_close_to. Examinez comment l’appel atteint mercantile.tile et comparez le résultat avec la gestion attendue de NaN dans l’issue ; le travail est terminé lorsque la requête n’échoue plus de manière inattendue pour cette entrée et que le comportement est couvert.