'module' object has no attribute 'boosted_trees'

User 1112 | 12/23/2014, 6:18:59 AM

I'm new to GraphLab and trying out "Using Gradient Boosted Trees to Predict Bike Sharing Demand."

I copy and pasted the exact code into Ipython Notebook and the file loads but I'm getting an error message that boosted_trees isn't a module in graphlab. Anybody know what I might be doing wrong?


User 6 | 12/23/2014, 6:23:53 AM

Apologies, function was renamed to boostedtreesclassifier, see here: http://graphlab.com/products/create/docs/generated/graphlab.boostedtreesclassifier.BoostedTreesClassifier.html

We will fix documentation.

User 1112 | 12/23/2014, 7:09:55 AM


Looks like numiterations is not recognized, is that equivalent to maxiterations?

Also 'objective' wasn't recognized either, does it need that? Or does it automatically work for regression?

User 14 | 12/23/2014, 7:15:40 AM

Yes. num_iterations has been renamed to max_iterations.

objective is no longer needed since we break the original boostedtrees into boostedtreesclassifier and boostedtrees_regression.

The blog post was written a while ago with API 0.9. You can find a ipython notebook version of the blog http://graphlab.com/learn/gallery/notebooks/kagglebikeshare_prediction.html where we try to update with the latest version GraphLab Create.

User 1112 | 12/23/2014, 7:18:35 AM

Ok great. Thanks so much!