Project Review

[jordanwillis]

Rubric Score

Criteria 1: Valid Python Code

• Score Level: 4/4
• Comment(s): Great job writing valid code, I didn't encounter any errors or issues!

Criteria 2: Implementation of Project Requirements

• Score Level: 3/4
• Comment(s): You did a nice job implementing the project requirements! I found a couple of small issues that I described below.

1. Don't forget that sentences can end in other characters beyond just periods in your get_average_sentence_length function (for example ? or !). Therefore, you should account for these in your logic as well like this.

def get_average_sentence_length(text):
    no_questions = text.replace('?', '.')
    no_exclaim = no_questions.replace('!', '.')
    sentences = no_exclaim.split('.')
    sentence_lengths = []

    for sentence in sentences:
        words = sentence.split()
        sentence_lengths.append(len(words))

    average = sum(sentence_lengths) / len(sentence_lengths)
    return average

2. Don't forget you need to divide by 3 in your find_text_similarity function before returning. Your current implementation is adding the 3 values together but not performing the division. The reason for the division is because we need to take an average of each of the different similarity scores.

Criteria 3: Software Architecture

• Score Level: 4/4
• Comment(s): Great work, you split everything up into proper classes and functions!

Criteria 4: Uses Python Language Features

• Score Level: 4/4
• Comment(s): Great work effectively using many python built-in functions and language features! I especially liked your very nice prepare_data implementation! I had not thought of that approach yet, plus one-liners that are easy to read and make sense are always slick.

Criteria 5: Produces Accurate Output

• Score Level: 3/4
• Comment(s): Congratulations, you found the correct murderer! Even though you identified the correct person your output was technically incorrect due to the division problem in your find_text_similarity. Here is what the correct answers were.

Lily Trebuchet 
0.553127686544467
Myrtle Beech 
0.13770600101747985
Gregg T. Fishy 
0.43826902657493294

If you think about it, we would not expect an answer to ever be greater than 1 because even if you compared the same text together, the best similarity score that they could have is 1 (e.g. 100%).

Overall Score: 18/20

Overall you did a fantastic job on this project and demonstrated that you learned a lot in this course! Keep up the great work!!

[larryb100]

Hi Jordan Thanks for the really positive feedback I really appreciate it!!! And thanks for pointing out that mistake with the value being greater than 1, I did notice it and it bothered me at the time but I felt a bit under pressure to get the assignment in as I am really short of time this week. It was a great project and the penny finally dropped (or at least got much closer to the ground) in terms of Classes and how they worked and why they were so useful so I feel really positive about that!! Thanks again Marcus

[jordanwillis]

No worries, that pressure can unfortunately get in the way sometimes! The good thing is that it was just an omission mistake (e.g. you forgot to add the division) and not an implementation or coding mistake! 👍