Databases Reference
In-Depth Information
four homes listed that do not contain an entry for number of bathrooms.
Hopefully, this implies “unknown” rather than “non-existent”.
The column “den” has 2,656 missing values while the cardinality is one.
Hovering over cardinality for den shows us that the single value is “Y”. In this
case, we may want to assume that when entering the data, the realtor did not
check the “den” box, because there was no den in the home. Therefore we may
want to assign “N” those missing values.
There are no missing values for the columns: city, daysOnMarket, latitude,
longitude, price, propertyType, SqFeet, state, street, yrBuilt, and zip.
Close the Summary Statistics window.
Right-click on the Homes dataset; select “Handle Missing Values”.
VisMiner requires that you specify a handling option for all columns with
missing data. The “OK” button is disabled (grayed) in this form until you have
specified handling options for all listed columns or until you decide to cancel
out of the operation.
Click on the “bathrooms” column. The handling options appear to the right
(Figure 3.2). Given that only 0.1% of the observations are missing a
Figure 3.2
Missing Values Handling Options
