Chap 4 suggestions

[salbert83]

Another excellent chapter! Some minor suggestions:

1. in spam_predict, replace [wrd for wrd in email_words if wrd in vocabulary] with intersect(email_words, model.vocabulary). Also, I think the code on the webpage has a typo (vocabulary instead of model.vocabulary).
2. Use log(probability) instead of probability to avoid numerical errors
3. In spam_filter_accuracy, why record predictions? It is unused and not returned.

Thank you.

[salbert83]

One more, replace
all_words_text = StringDocument(string([string(word, " ") for word in all_words]...))
with
all_words_text = StringDocument(join(all_words, " "))

Thanks

[salbert83]

Also, think you're missing the line to load the file, something like
raw_df = DataFrame(CSV.File(email_path))

Thanks

[pefontana]

Hi @salbert83 !
Thanks for the suggestions!
We are currently working on an updated of chapter 4, so this comments will help us a lot
You are right in:

• the raw_df = DataFrame(CSV.File(email_path)) line is missing
• We have some typos errors
• We are not using the predictions array

We are going to use the code suggestions!
In a few day we will upload the new chapter version and you will see the improvements.

Thank you very much!