I've long thought at the back of my mind that while the pipeline results offer a lot of filters attached to the sources, there isn't currently a good method to easily filter the measurements and then re-calculate the sources. After a request from @danieleUTS for exactly this kind of functionality, it's worth adding this.
A simple example would be if a user wanted to remove all the measurements that had a sibling from the measurements df and then use these to construct the sources.
This would mean bringing over the aggregate code from the pipeline, but shouldn't be too bad. Especially with Vaex this might be pretty quick.
A side note that vast-tools is the place for this, this is not really a pipeline issue as the job of the pipeline is to provide the data to manipulate further beyond its default metrics.
I've long thought at the back of my mind that while the pipeline results offer a lot of filters attached to the sources, there isn't currently a good method to easily filter the measurements and then re-calculate the sources. After a request from @danieleUTS for exactly this kind of functionality, it's worth adding this.
A simple example would be if a user wanted to remove all the measurements that had a sibling from the measurements df and then use these to construct the sources.
This would mean bringing over the aggregate code from the pipeline, but shouldn't be too bad. Especially with Vaex this might be pretty quick.
A side note that vast-tools is the place for this, this is not really a pipeline issue as the job of the pipeline is to provide the data to manipulate further beyond its default metrics.