szsctt / pwp-capstones Goto Github PK

Rubric Score

Criteria 1: Valid Python Code

• Score Level: 4
• Comment(s): The provided code runs without errors.

Criteria 2: Implementation of Project Requirements

• Score Level: 4
• Comment(s): You provide correct implementations for the required suite of functions within the scope of the assignment.

Criteria 3: Software Architecture

• Score Level: 3
• Comment(s): The code is organized in a reasonable way for the most part. I point out one place where reorganization is necessary in my second comment.

Criteria 4: Uses Python Language Features

• Score Level: 4
• Comment(s): You make good use of Python built ins to help you accomplish your work.

Criteria 5: Produces Accurate Output

• Score Level: 4
• Comment(s): Your output is well formatted and accurate.

Overall Score: 19/20

Please see my feedback for further comments.

Thanks for your submission. Overall, you did a good job. Most of this feedback are just recommendations.

1

As you're printing each author along with the average number of words in their note/intros, you write

print("Author: {}, average words per sentence: {}\n".format(person, str(get_average_sentence_length(names_texts[person]))))

Note that it is not necessary to cast the result of the call to get_average_sentence_length to a string. The following accomplishes the same thing

print("Author: {}, average words per sentence: {}\n".format(person, get_average_sentence_length(names_texts[person])))

2

It will be better to move sections Creating The Definition for Our Model and Creating our TextSample Instances to a later point in your notebook since they cannot be executed in sequence at the moment.

3

In your implementation of frequency_comparison, you first store the values of appearances and mutual appearances in a dictionary and then execute a sum operation over the dictionaries later. Although this works, it's unnecessary extra work. Since you see each value before storing it in the dictionary, simply keep a running sum of the appearances as you go along. This saves time. Here's a snippet of what the change could look like

appearances = 0
mutual_appearances = 0
# some code
if table1[element] > table2[element]:
    appearances += table1[element]
    mutual_appearances += table2[element]
else:
    appearances += table2[element]
    mutual_appearances += table1[element]
# the rest with similar changes

4

My final comment concerns the return statements in your get_average_sentence_length and frequency_comparison functions. Here they are

# get_average_sentence_length
return mean(num_words)

# frequency_comparison
sum(mutual_appearances.values()) / sum(appearances.values())

Although your code does not throw an error here, it can. You should make a habit of checking that a value is nonzero before dividing by it. In the first case, numpy will return nan if num_words is an empty list and in the second case, you will get a zero division exception. In both cases, your program will exit at some point with an unhandled exception. You can avoid these issues by checking before dividing.