Leveraging xarray

[mlee03]

Discussion about holding data as Xarray objects, for example, GridObj would hold the coordinates as a xarray Coordinate object. And regrid would hold field data as a xarray's DataArray.

[mlee03]

So, the new xgridobj would look something like slide 2 here

[fmalatino]

I think I may need permission to view the slide

[fmalatino]

Based on the slide should GridObj be changed to something like:

@dataclass
class GridObj:
    grid: xr.Dataset
    gridfile: str
    
    @classmethod
    def from_file(cls, filepath: str) -> "GridObj":
        with xr.open_dataset(filepath) as ds:
            return cls(grid = ds, gridfile = filepath)

    def write_out_grid(self, filepath: str):
        self.grid.to_netcdf(filepath)

and then aspects of the variables contained within the Dataset can be accessed via:

grid_obj = GridObj.from_file("/location/of/netcdf")

data_x = grid_obj.grid.x.values                     # <class 'numpy.ndarray'>

name = grid_obj.grid.x.attrs["standard_name"]       # <class 'str'>

[mlee03]

yes!

[fmalatino]

Ok I will make these changes and update the PR for GridObj