Mordred descriptor calculation issues

Hello. I am interested in calculating Mordred descriptors within the KNIME Analytics Platform. I have sufficiently linked the Anaconda environment and used the Python script node for Mordred calculations based on their Github documentation.

The node executes successfully without errors, but every descriptor has NaN value. This is not the case when I run Mordred outside of KNIME. Why would this be unsuccessful with KNIME, and is there any way to successfully run Mordred within KNIME?

Hi just a tip
you will get higher chance of support if you upload your workflow with data as well
br