Having preprocessed the dataset, let's do some sanity checking using descriptive statistics techniques. 

We can implement this as shown here: 

dfs.info()

The output of the preceding code is as follows:

<class 'pandas.core.frame.DataFrame'>
Int64Index: 37554 entries, 1 to 78442
Data columns (total 6 columns):
subject 37367 non-null object
from 37554 non-null object
date 37554 non-null datetime64[ns, UTC]
to 36882 non-null object
label 36962 non-null object
thread 37554 non-null object
dtypes: datetime64[ns, UTC](1), object(5)
memory usage: 2.0+ MB

We will learn more about descriptive statistics in Chapter 5, Descriptive Statistics. Note that there are 37,554 emails, with each email containing six columns—subject, from, date, to, label, and thread. Let's check the first few entries of the email dataset:

dfs.head(10)

The output of the preceding code is as follows:

Note that our dataframe so far contains six different columns. Take a look at the from field: it contains both the name and the email. For our analysis, we only need an email address. We can use a regular expression to refactor the column.