Ah, ok. We can work around this, but will take some crafting - do you have a link to share? I think that’ll be easier to work through.
Essentially your RG needs be a list of one of those fields in order for each cell to be 1 item in the list, and then you’ll search the database to grab the other values using the index #. So, if we were to work from what you started - a list of datevalues only - the other 2 text would be:
search for user_usage_from_API:first item’s expected_usage:item# current cell’s index
search for user_usage_from_API:first item’s usage:item# current cell’s index
Both of those searches need to have user_id = current user’s id so that we’re retrieving the correct data entry.
So those searches mean you’re doing a database wide search for the entry with the correct user_id - the first item that matches that - then drilling down to the list field and then drilling further down to the item in that list. Cell #1 will display items #1 for expected usage and usage, cell #2 will display items #2… and so on. Since the RG is set up to be a list of datevalues - those are easier and can just be “current cell’s date values”
This is clearly a tricky route to get these lists to display, so I’m wondering if it would make more sense to split out each of these items into their own entry? So for every new datevalue/expected_usage/usage > a new User_Usage_from_API gets created… all with user_id properly assigned. That way your RG can just be a list of User_Usage_from_API with 1 constraint… user_id = current user… and then it all splits up nicely for you. These fields would, in turn, need to be single values and not list values.