Following multiple issues coming from different cities where local tests are passing and the ones on Kitt are failing for some of the students.
According to this thread you can't read a pickle file created with a different pandas version using pickle.load. I replaced the method to pd.read_pickle which is meant to solve the issue.
Tested locally with my own notebooks, everything runs well but it would be great if you cross checked it before we update entire package for the Data program.
What would be the procedure after? Post on data-teachers to ask students to update the nbresult? Send an automated message to all running batches?
Following multiple issues coming from different cities where local tests are passing and the ones on Kitt are failing for some of the students.
According to this thread you can't read a pickle file created with a different pandas version using
pickle.load
. I replaced the method topd.read_pickle
which is meant to solve the issue.Tested locally with my own notebooks, everything runs well but it would be great if you cross checked it before we update entire package for the Data program.
What would be the procedure after? Post on
data-teachers
to ask students to update thenbresult
? Send an automated message to all running batches?