Group vs Table.Buffer

The following are my personal thoughts.

The first solution, Table.Buffer is not a panacea. The method in the article is effective but dangerous. Because it will perform very poorly when the table is large, and may even have a counterproductive effect.

The second solution seems to have problems. When the amount of data is large, you can try the following methods: 1. Use Table.Group to group and only get the required grouping columns and the latest date 2. Use Inner Join to merge queries The reason for doing this is that the efficiency of Table.Select is very worrying.

In addition, when the data becomes more and more, you may use incremental refresh. But my personal understanding is that Power Query is more like a pipeline, not a data pool, and so much data should not be cached and calculated. If the power query query is slow to execute, you may also try to put it in DAX for calculation.