Join us on July 16th as we explore real-world Reinforcement Learning Learn more

Why plugin dedicated environment only Python 27?

Level 2
Why plugin dedicated environment only Python 27?

As per question, why the plugins dedicated environment can be defined only using python 2.7 as interpreter?

 

thank you 

4 Replies
Dataiker
Dataiker

Hi @Seymour93 , you can change the required python version for the plugin in the 

code-env/python/desc.json:

"acceptedPythonInterpreters": ["PYTHON36"]

 

The valid values are:

PYTHON27, PYTHON34, PYTHON35, PYTHON36

 

Andrey
R&D Engineer @ Dataiku
Dataiker
Dataiker

@Seymour93 just to caveat what Andrey mentioned. The order of actions should be:

- Create plugin.

- Add code env component to plugin. (do not create code-env itself)

- Modify file as Andrey indicated.

- Create code env from main plugin page selecting the desired Python version.

Level 2

Hi @Liev  and @Andrey 

Could you please kindly explain where or how i can add code env component to H2O plugin?

I am using Dataiku free version, with docker on windows10. I installed H2O plugin. When I used H2O recipes to build model, it had error "Error Type". When I chose option "build missing packages and run recipes", there was no error but it did not save the model after training. Without a model, I could not go further to prediction step.

Thank you very much.

0 Kudos
Level 2

OK, I succeeded in creating python36 env for H2O plugin as @Andrey explained. But still no model was saved after training.

0 Kudos