personalized_model.evaluate ERROR

User 2742 | 12/6/2015, 7:33:26 PM

personalizedmodel = graphlab.itemsimilarityrecommender.create(traindata, userid='USERIDhash', itemid='COUPONIDhash', userdata= userinfo, itemdata= iteminfo) PROGRESS: Recsys training: model = itemsimilarity PROGRESS: Warning: Ignoring columns ITEMCOUNT, enSmallAreaName; PROGRESS: To use one of these as a target column, set target = <columnname> PROGRESS: and use a method that allows the use of a target. PROGRESS: Preparing data set. PROGRESS: Data has 82491 observations with 19076 users and 13231 items. PROGRESS: Data prepared in: 0.40839s PROGRESS: Computing item similarity statistics: PROGRESS: Computing most similar items for 13231 items: PROGRESS: +-----------------+-----------------+ PROGRESS: | Number of items | Elapsed Time | PROGRESS: +-----------------+-----------------+ PROGRESS: | 1000 | 2.59654 | PROGRESS: | 2000 | 2.65058 | PROGRESS: | 3000 | 2.69762 | PROGRESS: | 4000 | 2.73765 | PROGRESS: | 5000 | 2.77467 | PROGRESS: | 6000 | 2.80569 | PROGRESS: | 7000 | 2.86073 | PROGRESS: | 8000 | 2.93879 | PROGRESS: | 9000 | 3.00083 | PROGRESS: | 10000 | 3.06288 | PROGRESS: | 11000 | 3.13693 | PROGRESS: | 12000 | 3.20998 | PROGRESS: | 13000 | 3.34209 | PROGRESS: +-----------------+-----------------+ PROGRESS: Finished training in 3.71547s

personalizedmodel.evaluate(testdata)

[ERROR] Toolkit error: Assertion failed: (C:/jenkins/workspace/Dato-Dev-Continuous-Build-Win/src/unity/toolkits/recsys/models/itemcf.cpp:1006): usernumitems>0 [0 > 0]


ToolkitError Traceback (most recent call last) <ipython-input-21-145d692e22b5> in <module>() ----> 1 personalizedmodel.evaluate(testdata)

C:\Users\jithendra\AppData\Local\Dato\Dato Launcher\lib\site-packages\graphlab\toolkits_model_workflow.pyc in wrapper(model, args, **kwargs) 13 @wraps(f) 14 def wrapper(model, args, kwargs): ---> 15 result = f(model, *args, kwargs) 16 dataset_label = None 17 dataset = []

C:\Users\jithendra\AppData\Local\Dato\Dato Launcher\lib\site-packages\graphlab\toolkits\recommender\util.pyc in evaluate(self, dataset, metric, excludeknownforprecisionrecall, target, verbose, kwargs) 1631 excludeknown=excludeknownforprecision_recall, 1632 verbose=verbose, -> 1633 kwargs) 1634 ret.update(results) 1635 if verbose:

C:\Users\jithendra\AppData\Local\Dato\Dato Launcher\lib\site-packages\graphlab\toolkits\recommender\util.pyc in evaluateprecisionrecall(self, dataset, cutoffs, skipset, excludeknown, verbose, kwargs) 1456 excludeknown=excludeknown, 1457 verbose=verbose, -> 1458 kwargs) 1459 1460 precisionrecallbyuser = graphlab.recommender.util.precisionrecallby_user(dataset, recs, cutoffs)

C:\Users\jithendra\AppData\Local\Dato\Dato Launcher\lib\site-packages\graphlab\toolkits\recommender\util.pyc in recommend(self, users, k, exclude, items, newobservationdata, newuserdata, newitemdata, excludeknown, diversity, randomseed, verbose) 1363 'randomseed' : randomseed 1364 } -> 1365 response = graphlab.toolkits.main.run('recsys_recommend'Markdown`�I�M! ��7# ++����FYI: If you are using Anaconda and having problems with NumPyHello everyone,

I ran into an issue a few days ago and found out something that may be affecting many GraphLab users who use it with Anaconda on Windows. NumPy was unable to load, and consequently everything that requires it (Matplotlib etc).

Comments

User 954 | 12/7/2015, 7:11:45 PM

Hi There,

Can you send your issue to support@dato.com? That would be great if you can attach a sample dataset with a minimal script that regenerate the problem.

Regards, Emad