-
Notifications
You must be signed in to change notification settings - Fork 11
Open
Description
I have been experiencing some problems that seem to be related to pandas
version 1.1.0.
The problem occurs after loading any data with crowdtruth.load()
. For instance, I tried loading the data in the Multiple Choice Task - Person Type Annotation in Video
tutorial notebook:
data, config = crowdtruth.load(
file = "../data/person-video-multiple-choice.csv",
config = TestConfig()
)
After that, all rows in the 'units' data are duplicated. As an example:
data["units"]["input.subtitles"].unique()
>> array(["Italian astronaut samantha cristoforetti uploads a strange photograph she's taken"],
dtype=object)
Fixing the pandas
version to 1.0.5
or 0.25.3
in my virtual env solved the issue, so I think there's some changes in pandas 1.1.0
that are causing problems in loading the data.
Metadata
Metadata
Assignees
Labels
No labels