Quality flags

[thiagovmdon]

Hi @kratzert and @simonmoulds. I am wondering about how useful would be to retain, where available, quality flags from the original data? My idea would be to have an additional column (e.g., time, variable, flag), and somewhere we could point the users to where to find the meaning of the flags?

I do not see currently our task as to translate the flags and homogenize (for now), but perhaps it would be good to retain, when available?

What are your thoughts on that?

[kratzert]

I think I discussed this with @simonmoulds at one point. Yes, I think we would want to keep these columns but ultimately, also unify the quality maps somehow. Maybe just translate the quality flags to "all good", "something wrong", so that users know which timestamps to trust and which don't, but then we don't have to care about specifics. Because some countries have dozens of different flags for different types of impacts. But I think ultimately, this is beyond our scope to map these flags to common names.

We could also include all provided columns and then add one extra column, which is our attempt to provide a homogeneous structure.

[thiagovmdon]

I think I discussed this with @simonmoulds at one point. Yes, I think we would want to keep these columns but ultimately, also unify the quality maps somehow. Maybe just translate the quality flags to "all good", "something wrong", so that users know which timestamps to trust and which don't, but then we don't have to care about specifics. Because some countries have dozens of different flags for different types of impacts. But I think ultimately, this is beyond our scope to map these flags to common names.

We could also include all provided columns and then add one extra column, which is our attempt to provide a homogeneous structure.

Great. It does make sense. We can put the part of the "tentative homogenization" as a discussion point for our next meeting then. From now I think it is not that hard to adapt the ones that have such information and retain.