How does a measure replace a calculated column here?

[brs09j]

They say best practice is to always use a measure when you can. Lets say I have a calculated column that is

Calculated Column = Sales [Unit Price] * Sales [Quantity]

Measure = SUMX (Sales, Sales [Unit Price] * Sales [Quantity] )

So a calculated column will evaluate the formula and give you an individual answer for each new row in the new column.

The measure will evaluate the formula as well but will just sum up all of the results at the end to give you one answer.

How are these alike? Is this only for visualizations? Does this have to do with the default summarization feature? Let me know what I'm missing here.

Thanks.

[GilbertQ · Accepted solution]

Hi @brs09j

The reason for using measures is that due to all the data being stored in memory, when you use a measure it will not have to iterate over each and every row to complete the calculation. Or literally go through each row one at a time. So that is how it can perform things so quickly and easily.

In terms of your example a SUMX will still iterate over each and every row based on what a SUMX function is designed to do.

You would be better off creating a measure which would be:

Measure = sum('Sales'[Unit Price]) * sum('Sales'[Quantity])

This will give you the same result, but a lot quicker.