C1_W1 Wrong Predicted values (predict_tweet)

Hello, I am running my predict_tweet function, after implementing it l still get wrong values that are so close. Its predicting positives and negatives correctly but with wrong values. Where could l be missing in the code? Thank you for the assistance.

Hi Abiton_Padera,

Check your code carefully in all graded functions - the public tests do not capture everything.

If you cannot find a problem in your code, you can send me your notebook by personal message and I can have a look.

Hi @Abiton_Padera,

After looking at your code for your Ex 4, I’d say you are complicating things when they are very simple and straightforward. You are rewriting code.

For your exercise, you are writing a new code to extract features of a tweet, when you should be simply calling the extract_features function you implemented as Ex 3.

Similarly for prediction, you are rewriting code again when you should simply be using the sigmoid function you implemented as Ex 1.

Reuse the code as you are expected to. Read through all of the instructions before attempting any exercise. I promise you, everything is provided in the instructions, all you have to do is just follow them. If you rewrite code and don’t follow the instructions, you might end up passing the unit tests, but you could end up failing the autograder tests. So follow what is asked of you.


Also, please don’t try to change the utils.py files or unit test files in order for you code to work.

You are only expected to work in the assignment files, which are .ipynb, and not manipulate any other given file.