Benoit Bovy

Another advantage of Treescope is that is seems to only depend on numpy. For the interactive array visualization it injects vanilla JavaScript code via Python in a way that appears...

> Would it make sense to add it as an optional dependency and keep a "classical" fallback repr? If we go that route we could rather include https://github.com/benbovy/xarray-fancy-repr as an...

> which seems wrong simply based on the words: The left hand side indicates a number of nodes, the right hand side a number of cells. FWIW, a Fastscapelib grid...

In #6204 the `n_grid_nodes` variable was set only once in `initialize()` after calling `init_surface_mesh()` for the appropriate geometry. https://github.com/benbovy/aspect/blob/0a1a7f951bd878667d67b7fd0ed9996a91a58bde/source/mesh_deformation/fastscapecc.cc#L51-L101 I'm not sure why it has been changed since then. I...

> p is the slop exponent for multidirectonal flow, this is computed internally in fastscapelib c++ now, right @benbovy ? @Minerallo in Fastscapelib C++ this is a parameter of the...

xref https://github.com/sphinx-doc/sphinx/issues/11460#issuecomment-2404521778

In general I find [https://tutorial.xarray.dev/](https://tutorial.xarray.dev/) really well done, it would definitely deserve more exposure / discoverability / integration with the main docs!

Hi @DWesl, Your example is working with the `main` branch, it has been fixed in #6889. ```python # after #6889 arr2.unstack("index0") # # array([[[ 0, 1, 2, 3, 4], #...

@ACHMartin the issue is when you do `newds['z'] = stacked.z`. In the last versions of Xarray multi-index levels have each their own (real) coordinates, for consistency and clarity we soon...

I don't think it is a duplicate. I don't know what @ACHMartin's last minimal example is trying to do, but now the best way to assign a pandas multi-index to...