Join Our 5-Week ML/AI Engineer Interview Bootcamp 🚀 led by ML Tech Leads at FAANGs
44. Bag of words vectorization
GeneralTurn raw text into numbers using Bag-of-Words (BoW) vectorization, a simple NLP preprocessing step that counts how often each word appears. Given a fixed vocabulary, represent each sentence as a count vector where each entry corresponds to a vocabulary word.
Requirements
Implement the function
Rules:
- Tokenize each text by lowercasing and splitting on whitespace.
- For each text, count occurrences of each vocabulary token and output counts in the exact vocabulary order.
- Words not in
vocabularyare ignored. - Return the result as a NumPy array.
- Do not use any prebuilt vectorizers (e.g., scikit-learn).
- Use only NumPy and built-in Python libraries.
Example
Output:
| Argument | Type |
|---|---|
| texts | np.ndarray |
| vocabulary | np.ndarray |
| Return Name | Type |
|---|---|
| value | np.ndarray |
Constraints
No sklearn or prebuilt vectorizers
Whitespace split; lowercase tokens only
Return NumPy array count matrix
Hint 1
Lowercase each text and split on whitespace to get tokens; ignore everything else.
Hint 2
Precompute a dictionary {vocab_token: column_index} so you can update counts in O(1) per token.
Hint 3
Initialize a zero NumPy matrix of shape (len(texts), len(vocabulary)), then for each token found in the vocab, increment result[row, vocab_index[token]].
Roles
Companies
GeneralLevels
Tags
Python Editor