Training from train-spam-small.txt and train-ham-small.txt
Testing from test-spam-small.txt and test-ham-small.txt
number of emails 3 vs 2
entire vocab {'viagra', 'phil', 'the'}
entire vocab size is 3
spam words {'viagra': 3, 'phil': 1, 'the': 3}
ham words {'the': 2, 'phil': 2, 'viagra': 0}
Beginning tests.
Testing spam emails.
Test email 1
priors= 0.6 0.4
likelihoods= 0.064 0.046875
probs= 0.0384 0.018750000000000003
TEST 1 2/3 features true -3.260 -3.977 spam right
1 out of 1 classified correctly.
Testing ham emails.
Test email 1
priors= 0.6 0.4
likelihoods= 0.023999999999999987 0.046875
probs= 0.014399999999999991 0.018750000000000003
TEST 1 0/3 features true -4.241 -3.977 ham right
1 out of 1 classified correctly.
Total: 2/2 emails classified correctly.