Adjust dem fails when also interpolating onto grid

[TPCollings]

Hello again,

When using the tutorial script DEM.ipynb, I get an error when running the command to process the DEM with the mesh grid nodes, whilst also adjusting for coastlines. Example error is below.

I've found a work around, to adjust coastlines before including the interpolation onto the mesh, and then use the output from that to sample onto the mesh, with adjust_dem=False.

Cheers,

Tom

dem = pdem.Dem(**geometry, coastlines = ne_i, dem_source=dem.Dataset, grid_x = x, grid_y = y,adjust_dem=True)

INFO:pyposeidon.dem:dem done

INFO:pyposeidon.dem:.. interpolating on mesh ..

INFO:pyposeidon.dem:.. using elevation values ..

INFO:pyposeidon.dem:dem on grid done

INFO:pyposeidon.utils.fix:adjust dem

DEBUG:pyposeidon.utils.fix:compute water and land

DEBUG:pyposeidon.utils.fix:invoke pygeos

DEBUG:pyposeidon.utils.fix:find wet and dry masks

DEBUG:pyposeidon.utils.fix:fix wet points

DEBUG:pyposeidon.utils.fix:fix dry points

DEBUG:pyposeidon.utils.fix:assemble dataset

WARNING:pyposeidon.dem:Adjusting dem failed, keeping original values

---

ValueError Traceback (most recent call last)
Input In [77], in <cell line: 1>()
----> 1 dem = pdem.Dem(**geometry, coastlines = ne_i, dem_source=dem.Dataset, grid_x = x, grid_y = y,adjust_dem=True)

File ~/anaconda3/envs/pypos_main1/lib/python3.8/site-packages/pyposeidon/dem.py:79, in Dem.init(self, dem_source, **kwargs)
77 logger.info("Dem already adjusted\n")
78 else:
---> 79 self.adjust(coastline)

File ~/anaconda3/envs/pypos_main1/lib/python3.8/site-packages/pyposeidon/dem.py:91, in Dem.adjust(self, coastline, **kwargs)
89 if not check:
90 logger.warning("Adjusting dem failed, keeping original values\n")
---> 91 self.Dataset = self.Dataset.drop_vars("adjusted")

File ~/anaconda3/envs/pypos_main1/lib/python3.8/site-packages/xarray/core/dataset.py:4834, in Dataset.drop_vars(self, names, errors)
4832 names = set(names)
4833 if errors == "raise":
-> 4834 self._assert_all_in_dataset(names)
4836 # GH6505
4837 other_names = set()

File ~/anaconda3/envs/pypos_main1/lib/python3.8/site-packages/xarray/core/dataset.py:4801, in Dataset._assert_all_in_dataset(self, names, virtual_okay)
4799 bad_names -= self.virtual_variables
4800 if bad_names:
-> 4801 raise ValueError(
4802 "One or more of the specified variables "
4803 "cannot be found in this dataset"
4804 )

ValueError: One or more of the specified variables cannot be found in this dataset

[brey]

I am testing it on binder and it seems to work fine there.