Preserve last updated records

[jochendecraene]

Hi

In order to keep the last updated records in my report I found 2 techniques:

- one that uses table.buffer in combination with sort and delete double records

- one that uses Group By

Now I'm wondering what the advantages or disadvantages are between these two technique and if there are any better solutions?

Any thougts?

[ZhangKun · Accepted solution]

Table. Buffer is used to cache results. For large tables, Table The Buffer function may slow down the calculation.
All data needs to be traversed whether it is de duplication or grouping. But, if the data source supports query folding (query folding may be interrupted by other operations), grouping is excellent, because grouping operations will be performed on the server instead of locally. However, if you use Table Buffer function to cache data is undoubtedly a very bad decision.
In conclusion, if the data source supports query folding, you should use grouping. I personally feel that there is no difference between other data sources.

[Anonymous]

Hi @jochendecraene ,

We hope you found the information useful. If the above reply was helpful, please mark it as a solution. thanks.

Best Regards,
Gao
Community Support Team

If there is any post helps, then please consider Accept it as the solution  to help the other members find it more quickly.
If I misunderstand your needs or you still have problems on it, please feel free to let us know. Thanks a lot!

How to get your questions answered quickly --  How to provide sample data in the Power BI Forum

[Omid_Motamedise · Accepted solution]

It is all related to your data structure, if the number of records is not that much there is no difference, but if you are faced with too much data, and the number of groups that you want to extract the last record for them is less than around thousand go for Table.Group otherwise use Table.Buffer and Table.Sort technique

[ZhangKun · Accepted solution]

Table. Buffer is used to cache results. For large tables, Table The Buffer function may slow down the calculation.
All data needs to be traversed whether it is de duplication or grouping. But, if the data source supports query folding (query folding may be interrupted by other operations), grouping is excellent, because grouping operations will be performed on the server instead of locally. However, if you use Table Buffer function to cache data is undoubtedly a very bad decision.
In conclusion, if the data source supports query folding, you should use grouping. I personally feel that there is no difference between other data sources.