hi ,

I am getting below error ,i could not understand ,what mistake i did.

Thank you

I added several print statements to instrument my `train_naive_bayes`

function and hereâ€™s what I see when I run that test cell:

```
type(wordlist) <class 'list'>
V = 9165, len(wordlist) 11436
V: 9165, V_pos: 5804, V_neg: 5632, D: 8000, D_pos: 4000, D_neg: 4000, N_pos: 27547, N_neg: 27152
freq_pos for smile = 47
freq_neg for smile = 9
loglikelihood for smile = 1.5577981920239676
0.0
9165
```

The most common error on this function is simply to add one to `N_pos`

and `N_neg`

for each occurrence. The intent here is that you add the actual frequency each time by looking it up in the `freqs`

dictionary that was passed in as an argument.

Of course there are other moving parts here, e.g. check the order of operations on your computation of `p_w_pos`

and `p_w_neg`

.

Yes, your values do all agree with mine. Itâ€™s probably time to look at your code. We arenâ€™t supposed to do that in public on this thread, but we can do it by DM. I will send you a DM about how to do that.

Hello! could you explain to me the concept of N_pos and N_neg?

We have a dictionary called `freqs`

where the keys are `(word, sentiment)`

where `sentiment`

is either 1 for â€śpositiveâ€ť or 0 for â€śnegativeâ€ť. What that dictionary tells us is the total number of times a given word appears in a positive tweet and appears in a negative tweet.

In order to implement the Naive Bayes formula we need to know the total number of occurrences of unique words in positive tweets and negative tweets. That is what the numbers `N_pos`

and `N_neg`

represent: the total number of occurrences of positive words and negative words.

To compute those numbers we can loop over the keys in the `freqs`

dictionary. Note that some words appear in both positive and negative tweets, but not all words do.

I have same problem - all my values is same, but I got this error:

Wrong values for loglikelihood dictionary. Please check your implementation for the loglikelihood dictionary.

Wrong values for loglikelihood dictionary. Please check your implementation for the loglikelihood dictionary.

Wrong values for loglikelihood dictionary. Please check your implementation for the loglikelihood dictionary.

12 Tests passed

3 Tests failed

Please add similar print statements to those I added earlier on this thread and show the results you get. Are you sure everything is the same?

Something must be different if you are failing the tests. The tests have been in use for years by now, so we would know if there were problems with the tests.

The most common error here is just adding one for each occurrence instead of using the actual frequencies from the `freqs`

dictionary.

V: 9165, V_pos: 5804, V_neg: 5632, D: 8000, D_pos: 4000.0, D_neg: 4000.0, N_pos: 27547, N_neg: 27152

req_pos for smile = 47

freq_neg for smile = 9

loglikelihood for smile = 1.6529230243738393

Yes I see difference in loglikelyhood. I used just np.log(p_w_pos/p_w_neg) and got this value. I calculated the example manually and got your result. Ok. Thank you. I see the way but didnâ€™t understand how this happen

How did you get this value?

p_w_pos = (47+1)/(27547+5804)

p_w_neg = (9+1)/(27152+5632)

np.log(p_w_pos/p_w_neg)

My result is 1.5514687524908015

Could you mark where can be mistake?

Upd. I found

Great! Yes, you have not translated the math formula correctly. Glad to hear you found the issue.

1 Like

Hi i passed all test cells in programming assignment as shown below but when i submit the notebook i got 0 in two cells tests even it display on test cell function that i passed how to solve this

That is the next assignment, not Naive Bayes. But that syndrome usually means you have made the mistake of referencing the global variable `word_embeddings`

in your `get_country`

function.

this is the function i rewrite it more than one time and same error display

[**Images removed by moderator as it contains grader cell codes and posting codes is against community guidelines**]

Yes, you are making exactly the mistake that I described.

where is the mistake ?

I explained it earlier: it is a mistake to reference `word_embeddings`

directly. It works in the notebook, but the grader passes a different dictionary. It is passed to you as an argument to the function, so the correct way is to use the argument value which is `embeddings`

, right?

right i noticed it now thank you

1 Like