Zeb Nicholls

@maritsandstad I'll post in this issue once I've had a first go at updating the adapter docs!

I think the easiest solution here is to alter pyam's behaviour so the nan drop is only if the value column is nan, not any column. However will pull in...

@miturbide are you the best person to ask about this?

Awesome thanks!

Ok great thanks, solutions sound good

I read through #875, #764 and #753 yesterday and found these super helpful background resources. Unfortunately they made things seem very complicated and I couldn't see an obvious way forward....

@andrewgsavage has done much more work on this than me, but maybe these thoughts are helpful nonetheless... The big problem is with how pandas decides whether something is iterable or...

Thanks @shoyer. Whilst others haven't commented yet, for reference I think the most relevant Pandas issue is https://github.com/pandas-dev/pandas/issues/24702 but it's not the same as the one we're having here

I had a quick play to see what it would take to fix this. It looks like a simple change, but more experienced eyes would be very welcome! See draft...

Ok done in https://github.com/pandas-dev/pandas/issues/35131 and https://github.com/pandas-dev/pandas/pull/35127. I'm not sure if we need to ping anyone to get those moving?