Hi, I like to export (or copy/paste) a decition tree output table with each input's logworth value (or some wort of variable ranking). Is this possible to do?
When I do decision tree in interactive mode and try to split, a window pops up with all available input variables with -log(p) values calculated. Unfortunately, I couldn't copy this table and also couldn't find anywhere as permanant table.
Anyone know how I can output this window pop up in excel or text?
Thanks!
I'm not sure whether you are inquiring about the candidate split values for a particular split you are considering interactively or the final logworth values for the model you have already fit.
For the candidate split values at a particular node, you could do a screen capture but the dialog does not provide the capability to be saved to a spreadsheet. The values of the candidate splits at a given node are dependent on those splits made prior to the node in the tree so it would not make sense to persist these temporary values until the model has been fit.
After the model has been fit interactively and the interactive window is closed, you can click on
View --> Model --> Variable Importance
to surface these values for the variables in the fitted model, and with that table as the active window you can click on
File --> Save As...
to save the file with Variable Importance values as a SAS data set or as a .CSV file.
I hope this helps!
Doug
I'm not sure whether you are inquiring about the candidate split values for a particular split you are considering interactively or the final logworth values for the model you have already fit.
For the candidate split values at a particular node, you could do a screen capture but the dialog does not provide the capability to be saved to a spreadsheet. The values of the candidate splits at a given node are dependent on those splits made prior to the node in the tree so it would not make sense to persist these temporary values until the model has been fit.
After the model has been fit interactively and the interactive window is closed, you can click on
View --> Model --> Variable Importance
to surface these values for the variables in the fitted model, and with that table as the active window you can click on
File --> Save As...
to save the file with Variable Importance values as a SAS data set or as a .CSV file.
I hope this helps!
Doug
Build your skills. Make connections. Enjoy creative freedom. Maybe change the world. Registration is now open through August 30th. Visit the SAS Hackathon homepage.
Register today!Use this tutorial as a handy guide to weigh the pros and cons of these commonly used machine learning algorithms.
Find more tutorials on the SAS Users YouTube channel.