Forum Discussion
Difference between changing data type in query editor versus in data view
davidwsw Yes there is a difference. Any steps / transformations you apply at the query editor level will be iterated through each time you refresh / data is sourced from the source and so that will reflect through in Data view. Any changes you make in Data view doesn't add a step in the list of steps for query editor and so you don't see that change when you go to query editor.
In short, changes made in query editor are all performed on data when it comes from the source while in Data view those changes are applied once data is imported into power bi desktop.
ankitpatira What is the advantage of one over the other? Does it mean that doing it in data view results in better performance?
davidwsw Yes there is a difference. Any steps / transformations you apply at the query editor level will be iterated through each time you refresh / data is sourced from the source and so that will reflect through in Data view. Any changes you make in Data view doesn't add a step in the list of steps for query editor and so you don't see that change when you go to query editor.
In short, changes made in query editor are all performed on data when it comes from the source while in Data view those changes are applied once data is imported into power bi desktop.
ankitpatira What is the advantage of one over the other? Does it mean that doing it in data view results in better performance?
davidwsw I am not 100% sure but I think if you do lots of transformations via query editor it may take longer than you do in the desktop as each transformation via query editor results in a step that gets applied each time you refresh data / import data. For a change like data type change I don't think it should be a big issue and so doing it in either is fine. I would do it in Data view so that data coming from the source in power bi is as raw as it is in source.
Hello,
ankitpatira, you say keeping the data as raw as it is in source. Is it a guideline and/or a good pratice? I thought the reason data transformation exists is to use it as much as possible.
Any other advantages to use one over the other?
Thank you
