Description

Azalea Gui & Peter Lin
Part 1: Text generation, uniformly random model
Iteration word words word_frequencies
0 N/A [] []
1 ’Hello’ [’Hello’] [2]
2 ’Amy’ [’Hello’, ’Amy’] [2, 1]
3 ’was’ [’Hello’, ’Amy’, ’was’] [2, 1, 1]
4 ’here’ [’Hello’, ’Amy’, ’was’, ’here’] [2, 1, 1, 1]
1. (a)
(b) Including a specific example as the doctest’s expected output when a function is random isn’t a good ideabecause the function’s output will be different from the expected output each time it is executed. Since the doctest only verifies if the actual output matches the expected output, specifying a single random outcome as the expected output among many other possibilities will likely produce an error when running the test.
(c) For example, you can use words = {’Hello’: 1} as the words dictionary. In this case, generate text uniform(words, 5) has only one possible outcome, which is
’Hello Hello Hello Hello Hello’, so we can use that as our statement and expected output in the doctest.
2. Complete this part in the provided a3 part1.py starter file. Do not include your solution in this file.
Part 2: Text generation, One-Word Context Model
0. This question is not to be handed in.
1. One-word context model:
{
’Love’: [’is’, ’is’],
’is’: [’patient.’, ’kind.’, ’not’],
’patient.’: [’Love’],
’kind.’: [’It’],
’It’: [’does’, ’does’, ’is’],
’does’: [’not’, ’not’],
’not’: [’envy.’, ’boast.’, ’proud.’],
’envy.’: [’It’],
’boast.’: [’It’]
}
2. Complete this part in the provided a3 part2.py starter file. Do not include your solution in this file.
Part 3: Loops and Mutation Debugging Exercise
1. The test test star wars passed, and the tests test legally blonde and test transformers failed.
2. i. The test test legally blonde failed because of a mistake in the funtion clean text. Since string values are not mutable, and the function str.lower(str) does not mutate the string, str.lower(text) did not convert the words in the string to lower case but created a new string with lower-cased letters of the original string instead. However, since the result of str.lower(text), the new lower-cased string, wasn’t stored back into text, the string value in the variable text is not lower-cased. Since the non-lowercased string is used when processing the words, some of the words could not be matched to the VADER intensities dictionary, and they were not counted in the intensity calculation even though they should be counted.
ii. The test test transformers failed because of a mistake in the function count keywords. It should loop through the word list, find words that have VADER intensity data, and create a dictionary of the number of occurences of these words. When creating the dictionary, it uses an accumulator occurences so far. The keys of the dictionary represent these words and the values represent the number of times these words appear. When a new word word is found that isn’t in the accumulator, it should initialize occurences so far[word] to 1, and when a word word that’s already in the accumulator is found, it should add one to occurences so far[word]. The given code completed the first part (initializing the counts of new words to 1) correctly, but it didn’t add one when existing words are detected, so the returned result always reported one occurence for each word when some words actually appeared multiple times. Since the reported word occurences were incorrect, the calculated intensity were multiplied by potentially the wrong amount, which lead to the incorrect intensity.
3. The test test star wars passed because the review did not contain repeating or non-lower-cased words. The review text only contains three words that are in the small subset of the VADAR lexicon used in the program: magnificent , adventure , and succeeded. The problem in 2.i didn’t affect this text because these three words are all already lower-cased, and the problem in 2.ii didn’t affect this text because these three words all appeared only once in the text.
Part 4: Forest Fires
1. Complete this part in the provided a3 part4.py starter file. Do not include your solution in this file.
2. Complete this part in the provided a3 part4 tests.py starter file. Do not include your solution in this file.
3. a. The precipitation attribute is a float, and in Python floats are immutable. In calculate mr, if ever the precipitation parameter is changed, it will be a variable reassignment, which will change the id of precipitation. This does not change the id of wm.precipitation, which is why calculate mr can never mutate the precipitation attribute of wm.
b. The function wants to use the given temperature or −2.8, whichever is higher. If the wm.temperature attribute is below −2.8 and gets reassigned to −2.8, then it will mutate the wm object, which is unwanted. However, if it is first stored into a different variable temperature, then any change to temperature won’t affect wm.temperature, and therefore wm won’t get mutated, which is good.
c. The elements of a tuple themselves can be mutated, as stated in the question. However, they cannot bereassigned, the id’s of the elements will all stay the same, and elements cannot be added or removed, which makes tuples immutable.