Im trying to write a function called estimate_ngram_probabilities to estimate the conditional probabilities of N-grams, utilizing the ConditionalFreqDist and ConditionalProbDist classes along with a probability distribution such as MLEProbDist from the nltk.probability module. The function should take the following parameter:

• ngrams_list: A list of N-grams for which to estimate conditional probabilities.

The function should return:

• A ConditionalProbDist object representing the conditional probabilities of the input N-grams.

Using the function, I’m then trying to create two variables named bigram_prob_dist and fivegram_prob_dist. The bigram_prob_dist variable is created using a list of bigrams, and the fivegram_prob_dist is created using a list of fivegrams.

The problem that Im having is that I’m getting an error “ValueError: too many values to unpack (expected 2)” when trying to run the fivegram list through the function. The bigram list runs fine.

Here is my function so far:

def estimate_ngram_probabilities(ngrams_list):
    cfdist = ConditionalFreqDist(ngrams_list)
    cpdist = ConditionalProbDist(cfdist, MLEProbDist)
    
    return cpdist

The fivegram list that im trying to pass to the function is in this format:

[('<s>', '<s>', '<s>', '<s>', '1609'),
 ('<s>', '<s>', '<s>', '1609', 'sonnets'),
 ('<s>', '<s>', '1609', 'sonnets', 'william'),
 ('<s>', '1609', 'sonnets', 'william', 'shakespeare'),
 ('1609', 'sonnets', 'william', 'shakespeare', '1'),
 ('sonnets', 'william', 'shakespeare', '1', 'fairest'),
 ('william', 'shakespeare', '1', 'fairest', 'creatures'), ...