Determine the minimum numeric value when column contains null/string values

Emilia Fuks

What problem are you facing?

The "minimum" aggregate returns null() as the minimum if at least one of the values in a column are empty.

What impact does this problem have on you/your business?

This impacts data quality and returns incorrect results for aggregates when dealing with numeric values.

Do you have any existing workarounds? If so, please describe those.

Still testing a workaround but currently trying to filter off null() values before the group step and then joining them back on.

Do you have any suggestions to solve the problem? Feel free to add images if this helps.

Add a numeric minimum and a numeric maximum aggregate, which would ignore null() or invalid values and return the minimum/maximum numeric value.

Henry Simms

Note - these new "Numeric minimum" and "Numeric Maximum" aggregates in the Group step should operate like the existing "numeric" aggregates like "Sum" and "Average" by automatically excluding / ignoring non-numeric values. The documentation describes this already