Save R objects in Azure ML and use them in other experiment?

Mike Ubezzi 2,776 Reputation points
2020-05-13T00:14:50.08+00:00

"Is it possible to save R object in Azure storage (not a trained model) and then use it in a new ML experiment?

My specific example is:
Running PCA with R (""prcomp"") and then apply the transformation over new data.

It can be done within a single R script component with the following code:

prin_comp <- prcomp(dataset1, scale= TRUE)
dataset2<-predict(prin_comp, newdata=dataset2)

But I want to save the new object ""prin_comp"" and call it in a new ML experiment.

Is it possible?

(I know there is PCA component in Azure ML but I'm still asking on this approach)."

[Note: As we migrate from MSDN, this question has been posted by an Azure Cloud Engineer as a frequently asked question] Source: MSDN

Azure Machine Learning
Azure Machine Learning
An Azure machine learning service for building and deploying models.
2,719 questions
0 comments No comments
{count} votes

Accepted answer
  1. Rohit Mungi 801 Reputation points
    2020-05-13T09:40:22.827+00:00

    Thanks for reaching out. Currently, you cannot save an R object in one experiment and call it directly in another experiment. You need to export the data into your azure storage and import the data into another experiment using the export/import data modules. In Azure ML (Preview), you can create/register datasets to use across experiments.

    Hope this helps. Thanks.

    Source: Azure Documentation

    0 comments No comments

0 additional answers

Sort by: Most helpful