Auto-Complete: Building the Auto-Complete System

load_dotenv("posts/nlp/.env", override=True)

path = os.environ["TWITTER_AUTOCOMPLETE"]
with open(path) as reader:
    data = reader.read()

# UNQ_C11 (UNIQUE CELL IDENTIFIER, DO NOT EDIT)
# GRADED FUNCTION: suggest_a_word
def suggest_a_word(previous_tokens, n_gram_counts, n_plus1_gram_counts, vocabulary, k=1.0, start_with=None):
    """
    Get suggestion for the next word

    Args:
       previous_tokens: The sentence you input where each token is a word. Must have length > n 
       n_gram_counts: Dictionary of counts of (n+1)-grams
       n_plus1_gram_counts: Dictionary of counts of (n+1)-grams
       vocabulary: List of words
       k: positive constant, smoothing parameter
       start_with: If not None, specifies the first few letters of the next word

    Returns:
       A tuple of 
         - string of the most likely next word
         - corresponding probability
    """

    # length of previous words
    n = len(list(n_gram_counts.keys())[0]) 

    # From the words that the user already typed
    # get the most recent 'n' words as the previous n-gram
    previous_n_gram = previous_tokens[-n:]

    # Estimate the probabilities that each word in the vocabulary
    # is the next word,
    # given the previous n-gram, the dictionary of n-gram counts,
    # the dictionary of n plus 1 gram counts, and the smoothing constant
    probabilities = estimate_probabilities(previous_n_gram,
                                           n_gram_counts, n_plus1_gram_counts,
                                           vocabulary, k=k)

    # Initialize suggested word to None
    # This will be set to the word with highest probability
    suggestion = None

    # Initialize the highest word probability to 0
    # this will be set to the highest probability 
    # of all words to be suggested
    max_prob = 0

    ### START CODE HERE (Replace instances of 'None' with your code) ###

    # For each word and its probability in the probabilities dictionary:
    for word, prob in probabilities.items(): # complete this line

        # If the optional start_with string is set
        if start_with is not None: # complete this line

            # Check if the beginning of word does not match with the letters in 'start_with'
            if not word.startswith(start_with): # complete this line

                # if they don't match, skip this word (move onto the next word)
                continue # complete this line

        # Check if this word's probability
        # is greater than the current maximum probability
        if prob > max_prob: # complete this line

            # If so, save this word as the best suggestion (so far)
            suggestion = word

            # Save the new maximum probability
            max_prob = prob

    ### END CODE HERE

    return suggestion, max_prob

sentences = [['i', 'like', 'a', 'cat'],
             ['this', 'dog', 'is', 'like', 'a', 'cat']]
unique_words = list(set(sentences[0] + sentences[1]))

unigram_counts = NGrams(sentences, 1).counts
bigram_counts = NGrams(sentences, 2).counts

previous_tokens = ["i", "like"]
word, probability = suggest_a_word(previous_tokens, unigram_counts, bigram_counts, unique_words, k=1.0)
print(f"The previous words are 'i like',\n\tand the suggested word is `{word}` with a probability of {probability:.4f}")
expected_word, expected_probability = "a", 0.2727
expect(word).to(equal(expected_word))
expect(math.isclose(probability, expected_probability, abs_tol=1e-4)).to(be_true)
print()

# test your code when setting the starts_with
tmp_starts_with = 'c'
word, probability = suggest_a_word(previous_tokens, unigram_counts, bigram_counts, unique_words, k=1.0, start_with=tmp_starts_with)
print(f"The previous words are 'i like', the suggestion must start with `{tmp_starts_with}`\n\tand the suggested word is `{word}` with a probability of {probability:.4f}")

expected_word, expected_probability = "cat", 0.0909
expect(word).to(equal(expected_word))
expect(math.isclose(probability, expected_probability, abs_tol=1e-4)).to(be_true)

sentences = [['i', 'like', 'a', 'cat'],
             ['this', 'dog', 'is', 'like', 'a', 'cat']]
unique_words = list(set(sentences[0] + sentences[1]))

unigram_counts = NGrams(sentences, 1).counts
bigram_counts = NGrams(sentences, 2).counts
trigram_counts = NGrams(sentences, 3).counts
quadgram_counts = NGrams(sentences, 4).counts
qintgram_counts = NGrams(sentences, 5).counts

n_gram_counts_list = [unigram_counts, bigram_counts, trigram_counts, quadgram_counts, qintgram_counts]
previous_tokens = ["i", "like"]
tmp_suggest3 = get_suggestions(previous_tokens, n_gram_counts_list, unique_words, k=1.0)

print(f"The previous words are 'i like', the suggestions are:")
display(tmp_suggest3)