Using R to make predictions with multiple data sets

[pvail@unitedfcu.com]

I have a project where I use multiple secondary data sets with automated feature discovery and time aware modeling to build the model.  I need to use R to generate predictions and deploy the model.  How do use the API to send multiple data sets to the model to generate predictions?

In previous projects, I would be able to send a data frame and use Predict() to generate predictions but that was only one data set.  I now have three data sets with time aware modeling.  How do I set that up?

[taylor_larkin]

Hi pvail@unitedfcu.com ,

To expand on what my colleague @dalilaB mentioned, the functionality you're requesting is not yet in the R client. This is being worked on as we speak. Alternatives to this would be either specifying secondary_dataset_config_id in upload_dataset using the python client or using the public API endpoints directly.

Cheers,

Taylor

[pvail@unitedfcu.com]

Is this the Composable ML? Do you have a link for me to reference?

[dalilaB]

Here is the answer from one of my colleague
If you are on 2.18 a custom script using the DataRobotPOST helper method is advised. So you can still make predictions in a feature discovery project by heating the raw API

[pvail@unitedfcu.com]

2.18.1

[dalilaB]

Which version of Datarobot R Package  are you using ?