Plotly express not compatible with pandas NAType/pandas.NA/pd.NA values

[devmcp]

Plotly express functions fall over when given dataframes containing pd.NA values. I think these should have the same behaviour as None values.

Example:

# This works
px.scatter(pd.DataFrame([1, 2, 3, None, 5]))

# This results in "TypeError: Object of type NAType is not JSON serializable"
px.scatter(pd.DataFrame([1, 2, 3, None, 5]).convert_dtypes())

[nicolaskruchten]

Thanks for pointing that out! We'll try to get it fixed soon :)

[nicolaskruchten]

Note to self: will need to look at

plotly.py/packages/python/plotly/_plotly_utils/utils.py

Lines 168 to 178 in b34bd6d

	@staticmethod
	def encode_as_pandas(obj):
	"""Attempt to convert pandas.NaT"""
	pandas = get_module("pandas", should_load=False)
	if not pandas:
	raise NotEncodable
	if obj is pandas.NaT:
	return None
	else:
	raise NotEncodable

plotly.py/packages/python/plotly/plotly/io/_json.py

Lines 507 to 526 in b34bd6d

	if pd is not None:
	if obj is pd.NaT:
	return None
	elif isinstance(obj, (pd.Series, pd.DatetimeIndex)):
	if numpy_allowed and obj.dtype.kind in ("b", "i", "u", "f"):
	return np.ascontiguousarray(obj.values)
	elif obj.dtype.kind == "M":
	if isinstance(obj, pd.Series):
	dt_values = obj.dt.to_pydatetime().tolist()
	else: # DatetimeIndex
	dt_values = obj.to_pydatetime().tolist()
	if not datetime_allowed:
	# Note: We don't need to handle dropping timezones here because
	# numpy's datetime64 doesn't support them and pandas's tz_localize
	# above drops them.
	for i in range(len(dt_values)):
	dt_values[i] = dt_values[i].isoformat()
	return dt_values

[astafan8]

This simple addition worked for me when i used px.timeline #3336. Note that I'm a plotly newbie.

stolen from https://github.com/GoogleCloudPlatform/covid-19-open-data/pull/265/files :)

[avm19]

#3705 partially fixed this issue. The issue is fixed when Python's json is used, but the problem persists for orjson.

import orjson  # assert that orjson is available

plotly.io.json.config.default_engine = 'json'
px.scatter(pd.DataFrame([1, 2, 3, None, 5]).convert_dtypes()).to_json()  # works
plotly.io.json.config.default_engine = 'auto'  # or `orjson`
px.scatter(pd.DataFrame([1, 2, 3, None, 5]).convert_dtypes()).to_json()  # fails

[Coding-with-Adam]

thank you for taking a look at this @avm19 .

[gvwilson]

Hi - we are tidying up stale issues and PRs in Plotly's public repositories so that we can focus on things that are still important to our community. Since this one has been sitting for a while, I'm going to close it; if it is still a concern, please add a comment letting us know what recent version of our software you've checked it with so that I can reopen it and add it to our backlog. If you'd like to submit a PR, we'd be happy to prioritize a review, and if it's a request for tech support, please post in our community forum. Thank you - @gvwilson