Datasets with the BigML Dashboard

2.2 Categorical Fields

BigML creates one bin per label contained in a categorical field. Each bin contains the number of instances that have a specific label, e.g. the example shown in Figure 2.3 has six labels, therefore the histogram shows six bins, and 245 instances of this field are labeled as “Spain”.

\includegraphics[width=0.5\textwidth ]{images/categorical-histogram}
Figure 2.3 Example of histogram for categorical fields

Note: when BigML encounters binary formatted fields (all values 0 or 1), it treats them as categorical rather than numeric. You may override this default in the source configuration. (See the section Updating Field Types of the Sources with the BigML Dashboard [ 22 ] .)

BigML allows you to have up to 1,000 different labels in a categorical field.