Azure Machine Learning Studio SelectColumnsTransform - how to patch or set web service input parameter?

Question

The sentiment analysis sample at https://gallery.azure.ai/Collection/Twitter-Sentiment-Analysis-Collection-1 shows use of Filter Based Feature Selection in the training experiment, which is used to generate a SelectColumnsTransform to be saved and used in the predictive experiment, alongside the trained model. The article at https://learn.microsoft.com/en-us/azure/machine-learning/studio/create-models-and-endpoints-with-powershell explains how you can programmatically train multiple models on different datasets, save those models and create then patch multiple new endpoints, so that each can be used for scoring using a different model. The same technique can also be used to create and save multiple SelectColumnsTransform outputs, for feature selection specific to a given set of training data. However, the Patch-AmlWebServiceEndpoint does not appear to allow a SelectColumnsTransform in a scoring web service to be amended to use the relevant itransform saved during training. An 'EditableResourcesNotAvailable' message is returned, along with a list of resources that can be edited which includes models but not transformations. In addition, unlike (say) ImportData, a SelectColumnsTransform does not offer any parameters that can be exposed as web service parameters.

So, how is it possible to create multiple web service endpoints programmatically that each use different SelectColumnsTransform itransform blobs, such as for a document classification service where each endpoint is based on a different set of training data?

Any information much appreciated.

Answer 1

Never mind. I got rid of the SelectColumnsTransform altogether (departing from the example experiment), instead using a R script in the training experiment to save the names of the columns selected, then another R script in the predictive experiment to load those names and remove any other feature columns.