Numerical Computing with Python

Grouping by continuous variables

When grouping in pandas, you typically use columns with discrete repeating values. If there are no repeated values, then grouping would be pointless as there would only be one row per group. Continuous numeric columns typically have few repeated values and are generally not used to form groups. However, if we can transform columns with continuous values into a discrete column by placing each value into a bin, rounding them, or using some other mapping, then grouping with them makes sense.