Yu-Kun Qian

>I have this in xarrayutils. Would be happy to migrate and adapt if people think this is useful. This is extremely useful, as `xgcm` aim to **add an understanding of...

Thanks for your guys' code snippets. I agree that the metric terms are important and we should use finite volume method and therefore more general for various grids. So I...

> I am not sure this gives the desired result. This does weight the variable, diffs it and then unweights it again with the same type of metric. Since you...

Thanks @jbusecke. After reading the source code, I understand much better. So I rewrite the codes using `get_metric`, still with default BCs set initially: ```python def div(U, V, grid): #...

Hi @jbusecke, I've noticed the [issue](#313) raised by @dcherian. Glad this is in progress. > I have been using grid.transform extensively in my work and it works super well. >...

Looking forward to solving this issue. Tell me if I could do something. I am in the middle of some isopycnal analyses so maybe I could do some tests or...

Hi @jbusecke, > The next code iterations will remove the axis methods (on which autogenerate relies). We will likely remove/heavily refactor the autogenerate module (see https://github.com/xgcm/xgcm/issues/487). I am worrying about...

Hi @rabernat, I hope that we can calculate vorticity in a more elegent way with the help of `xgcm` or `grid_ufunc`. So I am expecting the **'perfect'** one would be:...

Thanks @rabernat. This is really a great package and help me simplify many calculations. That is why many of my codes rely on `xgcm`. The recent heavy-refactor, however, makes some...

Thanks @timothyas. No doubt that this is a **BIG** improvement! So I did not plan to roll my `xgcm` back to old version to save my codes. I just stand...