AssertionError Traceback (most recent call last)
<ipython-input-427-3044c83cef63> in <module>
1 from fastFM import bpr
2 fm = bpr.FMRecommender(n_iter=1000, init_stdev=0.1, rank=2, l2_reg_w=0.1, l2_reg_V=0.5)
----> 3 fm.fit(X_train, pairs_train)
4 y_pred = fm.predict(X_test)
/usr/local/lib/python3.5/dist-packages/fastFM/bpr.py in fit(self, X, pairs)
90 # check that pairs contain no real values
91 assert_array_equal(pairs, pairs.astype(np.int32))
---> 92 assert pairs.max() <= X.shape[1]
93 assert pairs.min() >= 0
94 self.w0_, self.w_, self.V_ = ffm.ffm_fit_sgd_bpr(self, X, pairs)
AssertionError:
The assert block should be making sure that the pairs are correct indices of X, by checking if the range of values in pairs fall within the number of samples of X
However, X.shape[1] represents the number of features, instead of number of samples
Should it be assert pairs.max() <= X.shape[0] instead?
for reference, the matrix X used in my code has a shape of (9276, 5992), i.e. more number of samples than number of features
When fitting BPR, I encountered this error:
The assert block should be making sure that the pairs are correct indices of
X
, by checking if the range of values inpairs
fall within the number of samples ofX
However,X.shape[1]
represents the number of features, instead of number of samples Should it beassert pairs.max() <= X.shape[0]
instead? for reference, the matrixX
used in my code has a shape of(9276, 5992)
, i.e. more number of samples than number of features