Here is a cool trick. Given a large data set, you may want to find out how often each object occurs. This is how you do it:

import pandas as pd

data = {'stocks': ['apple', 'microsoft', 'microsoft', 'apple', 'amazon', 'apple', 'apple', 'facebook', 'amazon']}
df = pd.DataFrame(data)

counts = df['stocks'].value_counts()
print(counts)

You should see this:

apple        4
microsoft    2
amazon       2
facebook     1
Name: stocks, dtype: int64

Pandas is amazing.