gb = df.groupby('category')
gb.groups
gb.get_group(1)
import ipdb;
ipdb.set_trace()
import os
if not os.path.exists('water-treatment.csv'):
!wget -O water-treatment.csv https://archive.ics.uci.edu/ml/machine-learning-databases/water-treatment/water-treatment.data
import IPython.display
IPython.display.Audio(url='http://soundeffect-lab.info/sound/button/mp3/decision8.mp3', autoplay=True)
with open("data/articles.csv") as f:
for i in range(100):
head = list(itertools.islice(f, 100))