`decode_psc` duplicates data

[JamesMcClung]

pscpy/src/pscpy/psc.py

Lines 116 to 121 in 8185aba

	data_vars = {}
	for var_name in ds:
	if var_name in field_to_component:
	for field, component in field_to_component[var_name].items(): # type: ignore[index]
	data_vars[field] = ds[var_name].isel({f"comp_{var_name}": component})
	ds = ds.assign(data_vars)

The code above creates 9 new variables (jx_ec, etc.) from the original jeh variable, but retains jeh. Unfortunately, the new variables aren't views; they are copies. To see this, note the "Size" field printed by the following demo:

import pscpy
import xarray

ds = xarray.load_dataset("path/to/pfd.000000000.bp")

print("Before decode:")
print(ds)

ds = pscpy.decode_psc(ds, ["e", "i"]) # or whatever species names there are

print("After decode")
print(ds)

When I run this on my data, the first print gives "Size: 18kB", and the second gives "Size: 37kB". When ds = ds.assign(data_vars) is removed, both have a size of 18 kB, but of course, then the latter dataset is missing the decoded variables. I think dropping the original jeh would fix this issue, but if jeh is required for some other reason, then this might be tricky to fix.