naive-bayes-movie-review-classifier.ipynb

{"cells":[{"metadata":{"_uuid":"8f2839f25d086af736a60e9eeb907d3b93b6e0e5","_cell_guid":"b1076dfc-b9ad-4769-8c92-a6c4dae69d19","trusted":true},"cell_type":"code","source":"# This Python 3 environment comes with many helpful analytics libraries installed\n# It is defined by the kaggle/python docker image: https://github.com/kaggle/docker-python\n# For example, here's several helpful packages to load in \n\nimport numpy as np # linear algebra\nimport pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)\nfrom scipy.sparse import csr_matrix, vstack\nfrom collections import Counter\nfrom tqdm import tqdm, trange\nfrom sklearn.datasets import load_svmlight_file\nfrom scipy.stats import norm\n# Input data files are available in the \"../input/\" directory.\n# For example, running this (by clicking run or pressing Shift+Enter) will list the files in the input directory\n\nimport os\nprint(os.listdir(\"../input/movie-ratings\"))\n\n# Any results you write to the current directory are saved as output.","execution_count":null,"outputs":[]},{"metadata":{"_cell_guid":"79c7e3d0-c299-4dcb-8224-4455121ee9b0","_uuid":"d629ff2d2480ee46fbb7e2d37f6b5fab8052498a","trusted":true},"cell_type":"code","source":"training_data, training_labels = load_svmlight_file('../input/movie-ratings/movie-ratings/data-splits/data.train')\ntesting_data, testing_labels = load_svmlight_file('../input/movie-ratings/movie-ratings/data-splits/data.test')\ntesting_data = csr_matrix((testing_data.data, testing_data.indices, testing_data.indptr), shape=(testing_data.shape[0], training_data.shape[1]))","execution_count":null,"outputs":[]},{"metadata":{"trusted":true,"_uuid":"1251c31a55706de89be3889b920f40f75ec6f20e"},"cell_type":"code","source":"print(training_data.shape, training_labels.shape)\nprint(testing_data.shape, testing_labels.shape)","execution_count":null,"outputs":[]},{"metadata":{"trusted":true,"_uuid":"105a7c77ecedd2089458c56a0661accc2b41f869"},"cell_type":"code","source":"gauss = norm()\ndef gaussian(mean, std_dev, value):\n    return np.clip(gauss.pdf((value - mean)/np.maximum(np.sqrt(std_dev), 1e-6)), a_min= 1e-6, a_max=None)","execution_count":null,"outputs":[]},{"metadata":{"trusted":true,"_uuid":"3609f58bf349c7b87265e8427a29e8029836140b"},"cell_type":"code","source":"class NB:\n    def __init__(self):\n        self.priors = None\n        self.probs =  None\n    \n    def fit(self, X, y):\n        y_counts = dict(Counter(y))\n        n = len(y)\n        self.priors = {key: np.log(value / n) for key, value in y_counts.items()}\n        \n        y1_idx = np.nonzero(y == 1)[0]\n        y0_idx = np.nonzero(y == 0)[0]\n        self.probs = []\n        for index, column in tqdm(enumerate(X.T)):\n            column = np.asarray(column.todense()).reshape(-1)\n            self.probs.append([None, None])\n            \n            # dealing with y = 0.\n            x_0 = column[y0_idx]\n            mean = np.mean(column[y0_idx])\n            std_dev = np.std(column[y0_idx])\n            self.probs[index][0] = ((mean, std_dev))\n            \n            # dealing with y = 1\n            x_1 = column[y1_idx]\n            mean = np.mean(column[y1_idx])\n            std_dev = np.std(column[y1_idx])\n            self.probs[index][1] = ((mean, std_dev))\n    \n    def predict(self, X):\n        predictions = np.zeros((X.shape[0], 2))\n        predictions[:,0] = self.priors[0]\n        predictions[:,1] = self.priors[1]\n        \n        X = X.tocsc()\n        for idx in tqdm(range(X.shape[1])):\n            col = np.asarray(X[:,idx].todense()).reshape(-1)\n            for possible_y in range(2):\n                mean, std_dev = self.probs[idx][possible_y]\n                predictions[:, possible_y] += np.log(gaussian(mean, std_dev, col))\n        return np.argmax(predictions, axis=1)\n    \n    def score(self, X, y):\n        assert X.shape[0] == y.shape[0]\n        return (self.predict(X) == testing_labels).mean()","execution_count":null,"outputs":[]},{"metadata":{"trusted":true,"_uuid":"1175051653fb97909c327e8022ebfd31e9d9c671","_kg_hide-output":true,"scrolled":false},"cell_type":"code","source":"nb = NB()\nnb.fit(training_data, training_labels)","execution_count":null,"outputs":[]},{"metadata":{"trusted":true,"_uuid":"dbebcb08c80a08fb81f2e4df6a4bdc4b57df3460"},"cell_type":"code","source":"nb.score(testing_data, testing_labels)","execution_count":null,"outputs":[]},{"metadata":{"trusted":true,"_uuid":"e90f5b156d52b77ac0b9dadc8d1cfd58fb9c7595"},"cell_type":"code","source":"(pred == testing_labels).mean()","execution_count":null,"outputs":[]},{"metadata":{"trusted":true,"_uuid":"9f62f00f186c374e0735c409978b152ee06c4e80"},"cell_type":"code","source":"eval_data, _ = load_svmlight_file('../input/movie-ratings/movie-ratings/data-splits/data.eval.anon')\neval_data = csr_matrix((eval_data.data, eval_data.indices, eval_data.indptr), shape=(eval_data.shape[0], training_data.shape[1]))\nsubmission_pred = nb.predict(eval_data)","execution_count":null,"outputs":[]},{"metadata":{"trusted":true,"_uuid":"8e72b1320d005ee1fe8a60e84b2de58b7abca859","scrolled":false},"cell_type":"code","source":"with open('submission.csv', 'w') as submission:\n    with open('../input/movie-ratings/movie-ratings/data-splits/data.eval.anon.id', 'r') as example_ids:\n        submission.write('example_id,label\\n')\n        for example_id, label in zip(example_ids, submission_pred):\n            submission.write('{},{}\\n'.format(example_id.strip(), int(label)))","execution_count":null,"outputs":[]}],"metadata":{"kernelspec":{"display_name":"Python 3","language":"python","name":"python3"},"language_info":{"name":"python","version":"3.6.6","mimetype":"text/x-python","codemirror_mode":{"name":"ipython","version":3},"pygments_lexer":"ipython3","nbconvert_exporter":"python","file_extension":".py"}},"nbformat":4,"nbformat_minor":1}