Exploring the Retailrocket dataset

Let's load the dataset and explore it to learn more about the data.

1. Set the path to the folder where we downloaded the data:

dsroot = os.path.join(os.path.expanduser('~'),
                      'datasets',
                      'kaggle-retailrocket')
os.listdir(dsroot)

2. Load the events.csv in a pandas DataFrame:

events = pd.read_csv(os.path.join(dsroot,'events.csv'))
print('Event data\n',events.head())

The events data has the five columns of timestamp, visitorid, event, itemid, and transactionid, as shown here:

Event data
        timestamp  visitorid event  itemid  transactionid
0  1433221332117     257597  view  355908            NaN
1  1433224214164     992329  view  248676            NaN
2  1433221999827     111016  view  318965            NaN
3  1433221955914     483717  view  253185            NaN
4  1433221337106     951259  view  367447            NaN

3. Print the unique items, users, and transactions:

print('Unique counts:',events.nunique())

 

We get the following...