Filtering based on row values

[da...@dataninjaa.com]

Im having trouble figuring out how to create a column that will list values greater or equal to the value in the current row.

for example I have Columns A,B,C

if C is not Null, return all cells from column A where column A's value is >= the value in column B

im fine with joining the array of results with some delimiter but im having difficulty figure out the right way to express this as a function.

thanks!

[Thad Guidry]

Have you seen the GREL Controls ?

And Variables ?

https://github.com/OpenRefine/OpenRefine/wiki/General-Refine-Expression-Language

We have example recipes here to help you learn the answer :)

https://github.com/OpenRefine/OpenRefine/wiki/Recipes

But let us know if you are still stuck after that and cannot figure it out.

Thad

https://www.linkedin.com/in/thadguidry/

--
You received this message because you are subscribed to the Google Groups "OpenRefine" group.
To unsubscribe from this group and stop receiving emails from it, send an email to openrefine+...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/openrefine/3a97e105-d9d1-40f0-ad20-bb93113a68ecn%40googlegroups.com.

[Tom Morris]

It's a little bit difficult to understand what you're asking. Perhaps you could demonstrate what you're looking for with a small (3x3?) example that's worked manually.

One thing to be aware of is that OpenRefine transforms work on a single row or the rows which are grouped together in a record. There aren't any operators which work on all cells in a column. It's unclear to me if that's one of the things that you're asking for.

Tom 

[da...@dataninjaa.com]

Thanks - yeah I had looked at those pages, I also started to try and convert the data set to one record so i could loop through the rows - but the dataset is too large - I decided to just switch to using a jupyter notebook with python. It is likely going to end up needing a large number of filtering and transformations based on specific data in certain rows against the whole dataset - so writing a script seems like a more suitable solution for this use case  in particular. 

I still use open refine quite a bit for various projects both for work and outside - thanks for your quick response!