medium_lstm_load_1.py

'''
Read reviews from a JSON-formatted file into an array.
'''
lines = [];
num_pos = 0; num_neg = 0; num_total = 75000;

with open('data/review.json', 'r') as f:
    for line in f:
        if (len(lines) >= (num_total * 2)):
            break;
        
        json_info = json.loads(line);
        
        if json_info['stars'] > 3:
            if num_pos > num_total:
                continue;
            num_pos = num_pos + 1;
        elif json_info['stars'] < 3:
            if num_neg > num_total:
                continue;
            num_neg = num_neg + 1;
        else:
            continue;
            
        lines.append(json.loads(line));

'''
Separate line data into reviews and labels
'''
reviews = [line['text'] for line in lines];

stars = [line['stars'] for line in lines];
labels = ['1' if star > 3 else '0' for star in stars];