Saving a model build in R or Python in the repository for re-use

KostasBonikosKostasBonikos Member Posts: 25 Maven
edited November 2018 in Knowledge Base

Typically in Rapidminer, we can save the resulting models from algorithms by using a simple "Store" operator.

 

We can then double-click those models and see their contents and we can also drop them into a new process and use them to score new data.

 

We can do this also when these models have been built in R or Python with the only difference that we cannot see inside the model artifacts. For example, if we drop and double-click a model build in Python or R, we get a fileobject without much description.

However, the model is there and can be used to score data without re-training.

 

See the example below:

 

save python model 1.png

Here, we have a gradient boosted trees model built in python and saved to the repository.

 

When trying to preview it, we get this:save python model 2.png

 

However, if we build another process where we bring the model artifact in and apply it to data, it works!

 

save python model 3.png

 

FYI, the code used for scoring was:

save python model 4.png

 

 

Tagged:

Comments

  • Rahul_MRahul_M Member Posts: 4 Contributor I
    What was the operator used to build the Apply Model (3) ? I haven't found an operator capable of accepting Python model as input
  • MartinLiebigMartinLiebig Administrator, Moderator, Employee, RapidMiner Certified Analyst, RapidMiner Certified Expert, University Professor Posts: 3,503 RM Data Scientist
    Hey @Rahul_M ,
    this article is a bit outdated. Please check the python learner operator,it covers this tasks in an easier way.

    Cheers,
    Martin
    - Sr. Director Data Solutions, Altair RapidMiner -
    Dortmund, Germany
Sign In or Register to comment.