def test_rmse(self):
eval = Evaluator()
model = ConstantModel(constant=3.0)
testing = pd.DataFrame([{'user':10, 'item':100, 'rating':3},
{'user':10,'item':110,'rating':1},
{'user':12,'item':100,'rating':4},
])
rmse = eval.evaluate_model_rmse(model, testing)
self.assertEqual(sqrt((0+4+1)/3.0), rmse)
eval.close()
sep="::", header=None, names=['user', 'item', 'rating', 'date', 'title'])
print df.head()
#tell me what models we want to evaluate
models = [
RandomModel(),
PopularityOkapi(host='54.72.18.118',
username='******',
#okapi_jar_dir='/Users/linas/devel/okapi/target/',
#okapi_jar_base_name='okapi-0.3.2-SNAPSHOT-jar-with-dependencies.jar',
public_key_path='/Users/linas/.ssh/hack-okapi.pem'
),
Popularity(normalize=False),
BPROkapi(host='54.72.18.118',
username='******',
#okapi_jar_dir='/Users/linas/devel/okapi/target/',
#okapi_jar_base_name='okapi-0.3.2-SNAPSHOT-jar-with-dependencies.jar',
public_key_path='/Users/linas/.ssh/hack-okapi.pem'
)
]
#setup the environment
evaluator = Evaluator()
for m in models:
m.fit(df)
print m.getName().ljust(50),
print evaluator.evaluate_model(m, df)
evaluator.close()#need this call to clean up the worker processes