Error in Knime python integration

[lmonno]

Hi, I have a python script that run regularly when I execute it in the Python Source node “execute script”.
The problem is when I run it in the workflow I get this error:

ERROR PythonKernel                    LookupError: No serializer extension having the id or processing python type "pandas._libs.tslibs.timestamps.Timestamp" could be found.
ERROR Python Source        0:306      Execute failed: An exception occured while running the python kernel. See console and log for details.

I have KNIME 3.7, python 3.6
thank you very much

[MarcelW]

Hi @lmonno,

Errors like this may be caused by pandas versions that are not compatbile with KNIME. What version of pandas are you using? Could you try to replace it with pandas 0.23 and see if the error vanishes?

Version requirements and recommendations are also documented in our Python installation guide. There, it is also described how environments that follow these recommendations can easily be created within KNIME.

Marcel

[lmonno]

Thank you Marcel for the answer.
I followed the reccomendation in your guide for versione 3.7 (here) and I created a yml file with pandas 0.23.
What I find strange is that I can run code without problems in the node (evaluate code) but not in the workflow.
Could be a format problem of my dataframe?

[lmonno]

I solved changing the format of the date with

my_table[‘my_date’]=my_table[‘my_date’].dt.strftime(‘%Y-%m-%d’)

Thanks anyway