WendyCzika Tracker

Re: SAS Miner - how to do model comparison without losing interactive decision tree settings

WendyCzika — Mon, 20 Nov 2023 16:21:58 GMT

The only thing I can think of is to check that the Input Data node at the start of the flow containing the Decision Tree node does not have the "Rerun" property set to Yes.

Re: SAS Miner - how to do model comparison without losing interactive decision tree settings

WendyCzika — Mon, 20 Nov 2023 13:19:03 GMT

I wouldn't think it would wipe out your interactive changes unless something is triggering the Decision Tree node to re-run, but you can set the Use Frozen Tree property to Yes to keep your interactive changes when it re-runs. Hope that helps!

Re: error 1012 will running neural network classification

WendyCzika — Wed, 27 Sep 2023 18:25:47 GMT

Please reach out to SAS Technical Support for help with this error

https://support.sas.com/en/technical-support.html

Re: Deploying/using machine learning model in SAS viya on new dataset to generate predictions

WendyCzika — Thu, 07 Sep 2023 12:40:58 GMT

Yes, the 1st, 3rd, and 4th bullets are all ways of scoring new data where the target is unknown. Hope that helps!

Re: Deploying/using machine learning model in SAS viya on new dataset to generate predictions

WendyCzika — Tue, 05 Sep 2023 13:58:23 GMT

There are several ways you can do this, depending on what exactly your needs are.

You can use the Score Data node
If you know the target in the data you want to score and also want to assess it, you can select Score Holdout Data from the Pipeline Comparison tab.
You can download score code from the Pipeline Comparison tab
You can register your model to Model Manager, which can be used to deploy and monitor models.

The last 3 bullets are covered here in the documentation: https://go.documentation.sas.com/doc/en/capcdc/v_020/vdmmlcdc/vdmmlug/p1ngioiguc5mmtn1e1wh8l5cew1r.htm

Re: Lesson 1.2 Demo Running Pipeline Issue for Starters Template

WendyCzika — Fri, 17 Feb 2023 20:13:30 GMT

Can you check that the variable you have assigned the role of Target on the Data tab has a Level other than Interval? If it's an interval target, you need to use Linear Regression instead of Logistic Regression. If that's not the issue, you can right-click on the Logistic Regression node, select Log, and see if there is a more informative error message in there.

Hope that helps,

Wendy

Re: What does it mean by setting the data role to transaction?

WendyCzika — Fri, 04 Nov 2022 14:11:49 GMT

In transaction data used by the Association, Market Basket, Path Analysis, and Time Series nodes in EM, each observation represents a product that a customer has purchased, so multiple observations per customer (which is the ID variable), which is different than the typical structure of data used for predictive modeling.

Re: Encoding categorical features in dataset

WendyCzika — Fri, 09 Sep 2022 16:10:43 GMT

I think so - I know that's true for other actions.

Re: Encoding categorical features in dataset

WendyCzika — Fri, 09 Sep 2022 15:29:03 GMT

Yes I think you can use the nominals= argument in the dlTrain action to indicate which inputs/target are categorical.

Re: Prediction After Random Forest

WendyCzika — Wed, 10 Aug 2022 19:27:50 GMT

You should be able to use the SCORE statement to create a data set with the predicted values for the training data: https://go.documentation.sas.com/doc/en/emhpprcref/15.2/emhpprcref_hpforest_syntax08.htm

Or you can use the SAVE statement to create an analytic store ("astore") file that allows you to score new data using that generated file in HP4SCORE: https://go.documentation.sas.com/doc/en/emhpprcref/15.2/emhpprcref_hpforest_syntax07.htm

and https://go.documentation.sas.com/doc/en/emhpprcref/15.2/emhpprcref_hp4score_toc.htm

Hope that helps.

Re: SPK Import Error for Model Manager

WendyCzika — Fri, 24 Jun 2022 15:36:54 GMT

Also for future issues in Model Manager, the Decisioning Community might be better suited to help you: https://communities.sas.com/t5/Decisioning/bd-p/decisioning

Re: autoneural.INVALIDARGUMENT_err

WendyCzika — Fri, 24 Jun 2022 15:29:24 GMT

To facilitate getting the information we need in order to debug this problem, can you please work with SAS Technical Support for this issue? https://support.sas.com/en/technical-support.html

Thank you!

Re: Missing value of False Negative in Classification Table - Gradient Boosting Model - SAS EM

WendyCzika — Mon, 18 Apr 2022 18:58:25 GMT

You might find some helpful tips for improving the gradient boosting model here:

https://communities.sas.com/t5/SAS-Communities-Library/Quick-tips-for-setting-your-Gradient-Boosting-node-properties-in/ta-p/513600

Re: How to import SAS diagram in SAS Enterprise Miner into SAS Model Studio

WendyCzika — Tue, 05 Apr 2022 12:27:01 GMT

A couple other things that could cause differences:

the metadata: make sure it is the same in both EM and Model Studio
if the data is big in Model Studio, the Batch Code node is sampling by default. Might have to make adjustments to that.

Another option would be to use the Score Code Import node in Model Studio after using a Score node in EM to accumulate all the score code from a flow.

Re: How to score data using a modelStore model with code?

WendyCzika — Fri, 01 Apr 2022 17:02:22 GMT

I think you want to do this:

libname mycas cas caslib=ModelStore;

Then use mycas where you have caslib and ModelStore in the proc gradboost/astore calls.

Re: How to import SAS diagram in SAS Enterprise Miner into SAS Model Studio

WendyCzika — Fri, 01 Apr 2022 12:50:57 GMT

You could be getting different results due to the partitioning of the data. Did you partition in EM?

Re: Import a xml into SAS Visual Data Mining and Machine Learning

WendyCzika — Thu, 31 Mar 2022 19:11:48 GMT

My guess is that it's related to partitioning of the data. In EM, did you partition your data?

Re: Cross validation autotuning and separate test set in Model Studio

WendyCzika — Mon, 14 Feb 2022 13:32:04 GMT

Do you know what version of SAS Viya you are on? It should be working in Viya 3.5 and on.

Re: Cross validation autotuning and separate test set in Model Studio

WendyCzika — Fri, 11 Feb 2022 19:00:50 GMT

Actually that doc isn't completely correct. If you only create training and test partitions (set the validation percentage to 0), then if you select K-fold cross validation for the Validation method, it will actually use that. So give that a try, and hope that helps!

Re: GBM model into Model Manager via PMML file

WendyCzika — Fri, 27 Aug 2021 19:50:59 GMT

Looking here under Supported PMML Models (almost at the end), I don't see GBM listed:

https://go.documentation.sas.com/api/docsets/mdlmgrug/15.2/content/mdlmgrug.pdf