Moodle Analytics – Part 2. – Setting up the Python Machine Learning Backend in Moodle

Gábor Zsolt NagyAugust 23, 2019Moodle9 comments

Once the Python Machine Learning Backend has been installed on the server, it is time to change the Moodle settings for better analytics.

Here are the steps.

Step 1.

Log in to Moodle with Site administrator role.
Navigate to “Site administration / Server / System paths”.
Set Path to Python.
Click on “Save changes” button.

Moodle Site administration / Server / System paths / Path to Python — Path to Python to set

Step 2.

Now we need to set the analytics on the main settings page. Unfortunately for every analytics model the settings need to be changed as well.

Navigate to “Site administration / Analytics / Analytics settings”.
Set Default predictions processor to Python machine learning backend.
Navigate to “Site administration / Analytics / Analytics models”.
Select “Edit” from the “Actions” dropdown menu for each model.
Set Predictions processor to Python machine learning backend.

Moodle Analytics settings, Default predictions processor — Analytics settings to set

Analytics model, Prediction processor setting — Changing the Prediction processor to Python machine learning backend

That’s all folks.

In the next Part the Tensorboard evaluation will be introduced.

Cheers,

Gábor Zsolt Nagy

9 thoughts to “Moodle Analytics – Part 2. – Setting up the Python Machine Learning Backend in Moodle”

Purwoko Haryadi Santoso October 24, 2022 at 09:39

I have followed the Previous Post (Install the moodlemlbackend on the server). However, I found issue in Step 1. It said “The selected predictions processor is not ready: The moodlemlbackend Python package is not installed or there is a problem with it. Please execute “/home/*********/virtualenv/analytics/3.7/bin/python -m moodlemlbackend.version” from command line interface for more info.”

Any solution? Thank you.

Reply
- Gábor Zsolt Nagy October 24, 2022 at 13:08
  
  As I can see, you installed Python 3.7. Please check if it is compatible with the latest moodlebackend.
  https://github.com/moodlehq/moodle-mlbackend-python/blob/master/requirements.txt
  What Moodle version are you using?
  
  Reply
  - Purwoko Haryadi Santoso October 25, 2022 at 09:15
    
    My moodle is on Version 4.0.3 (Build: 20220822).
    
    Reply
    - Gábor Zsolt Nagy October 27, 2022 at 10:24
      
      As mentioned in my article for me only Python 3.4 worked properly.
      Have you tried that Python version?
      
      Reply
      - Purwoko Haryadi Santoso October 29, 2022 at 17:08
        
        Unfortunately, there is no option to choose Python 3.4 in my cpanel version from Indonesian shared hosting. Instead, it recommends minimum 3.6.15 up to 3.10.4.
        
        Can we communicate via email to consult in more detail? Thank you.
      - Gábor Zsolt Nagy November 1, 2022 at 09:39
        
        I am going to check the Python versions and will let you by mail.
        Gábor
      - Gábor Zsolt Nagy November 3, 2022 at 22:29
        
        Successfully installed moodlemlbackend 3.0.5 with Python 3.6. Check the new article:
        Moodle Analytics – Part 1. – Setting up the Python Machine Learning Backend in cPanel – Reloaded
Purwoko Haryadi Santoso November 6, 2022 at 14:59

Thanks for the recent updated post. Actually, my issue occurs in part 2 of your tutorial as described in the current page. I have successfully installed the moodlemlbackend.

Maybe, you can read my post in the Moodle forum in https://moodle.org/mod/forum/discuss.php?d=426515

Reply
- Gábor Zsolt Nagy November 6, 2022 at 22:56
  
  I sent an email to you 3 days ago. You can reply to that address.
  
  I have never faced this problem.
  What I suggest is to change the debugging setting to developer in Moodle Site administration / Development / Debugging, and check the log.
  
  Reply