importing_data.py

%matplotlib inline

import os
import tarfile
import urllib
import shutil
import json
import random
import numpy as np
import boto3
import sagemaker

from tqdm import tqdm
from sagemaker.amazon.amazon_estimator import get_image_uri
from matplotlib import pyplot as plt
from xml.etree import ElementTree as ET
from PIL import Image, ImageDraw, ImageFont

urls = ['http://www.robots.ox.ac.uk/~vgg/data/pets/data/images.tar.gz',
        'http://www.robots.ox.ac.uk/~vgg/data/pets/data/annotations.tar.gz']

def download_and_extract(data_dir, download_dir):
    for url in urls:
        target_file = url.split('/')[-1]
        if target_file not in os.listdir(download_dir):
            print('Downloading', url)
            urllib.request.urlretrieve(url, os.path.join(download_dir, target_file))
            tf = tarfile.open(url.split('/')[-1])
            tf.extractall(data_dir)
        else:
            print('Already downloaded', url)

if not os.path.isdir('data'):
    os.mkdir('data')

download_and_extract('data', '.')