Skip to content

Commit a35d65b

Browse files
rbeucherrbeucher
committed
Restructure
1 parent 2a9f247 commit a35d65b

File tree

10 files changed

+1022
-1768
lines changed

10 files changed

+1022
-1768
lines changed

‎.gitignore‎

Lines changed: 1 addition & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -3,3 +3,4 @@ docs/build
33
src/access_mopper.egg-info
44
# pixi environments
55
.pixi
6+
notebooks/.ipynb_checkpoints

‎notebooks/GettingStartedOcean.ipynb‎

Lines changed: 0 additions & 883 deletions
This file was deleted.

‎notebooks/Getting_started.ipynb‎

Lines changed: 278 additions & 167 deletions
Large diffs are not rendered by default.

‎src/access_mopper/__init__.py‎

Lines changed: 1 addition & 6 deletions
Original file line numberDiff line numberDiff line change
@@ -1,11 +1,6 @@
11
from . import _version
22
from ._config import _creator
3-
from .cmip6_cmoriser import (
4-
ACCESS_ESM_CMORiser,
5-
CMIP6_Atmosphere_CMORiser,
6-
CMIP6_CMORiser,
7-
CMIP6_Ocean_CMORiser,
8-
)
3+
from .driver import ACCESS_ESM_CMORiser
94

105
__version__ = _version.get_versions()["version"]
116

‎src/access_mopper/_version.py‎

Lines changed: 1 addition & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -12,12 +12,12 @@
1212
"""Git implementation of _version.py."""
1313

1414
import errno
15+
import functools
1516
import os
1617
import re
1718
import subprocess
1819
import sys
1920
from typing import Any, Callable, Dict, List, Optional, Tuple
20-
import functools
2121

2222

2323
def get_keywords() -> Dict[str, str]:

‎src/access_mopper/atmosphere.py‎

Lines changed: 216 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,216 @@
1+
import numpy as np
2+
import xarray as xr
3+
4+
from access_mopper.base import CMIP6_CMORiser
5+
from access_mopper.derivations import custom_functions, evaluate_expression
6+
7+
8+
class CMIP6_Atmosphere_CMORiser(CMIP6_CMORiser):
9+
"""
10+
Handles CMORisation of NetCDF datasets using CMIP6 metadata (Atmosphere/Land).
11+
"""
12+
13+
def select_and_process_variables(self):
14+
# Find all required bounds variables
15+
bnds_required = []
16+
bounds_rename_map = {}
17+
for dim, v in self.vocab.axes.items():
18+
if v.get("must_have_bounds") == "yes":
19+
# Find the input dimension name that maps to this output name
20+
input_dim = None
21+
for k, val in self.mapping[self.cmor_name]["dimensions"].items():
22+
if val == v["out_name"]:
23+
input_dim = k
24+
break
25+
if input_dim is None:
26+
raise KeyError(
27+
f"Can't find input dimension mapping for output dimension '{v['out_name']}'."
28+
)
29+
bnds_var = input_dim + "_bnds"
30+
bounds_rename_map[bnds_var] = v["out_name"] + "_bnds"
31+
bnds_required.append(bnds_var)
32+
33+
# Select input variables
34+
input_vars = self.mapping[self.cmor_name]["model_variables"]
35+
calc = self.mapping[self.cmor_name]["calculation"]
36+
37+
required_vars = set(input_vars + bnds_required)
38+
self.load_dataset(required_vars=required_vars)
39+
40+
# Handle the calculation type
41+
if calc["type"] == "direct":
42+
# If the calculation is direct, just rename the variable
43+
self.ds = self.ds.rename({input_vars[0]: self.cmor_name})
44+
elif calc["type"] == "formula":
45+
# If the calculation is a formula, evaluate it
46+
context = {var: self.ds[var] for var in input_vars}
47+
context.update(custom_functions)
48+
self.ds[self.cmor_name] = evaluate_expression(calc, context)
49+
# Drop the original input variables, except the CMOR variable and keep bounds
50+
self.ds = self.ds.drop_vars(
51+
[
52+
var
53+
for var in input_vars
54+
if var != self.cmor_name and var not in bnds_required
55+
],
56+
errors="ignore",
57+
)
58+
else:
59+
raise ValueError(f"Unsupported calculation type: {calc['type']}")
60+
61+
# Rename dimensions according to the CMOR vocabulary
62+
dim_rename = self.mapping[self.cmor_name]["dimensions"]
63+
dims_to_rename = {k: v for k, v in dim_rename.items() if k in self.ds.dims}
64+
self.ds = self.ds.rename(dims_to_rename)
65+
66+
# Also rename coordinates if needed
67+
coords_to_rename = {k: v for k, v in dim_rename.items() if k in self.ds.coords}
68+
if coords_to_rename:
69+
self.ds = self.ds.rename(coords_to_rename)
70+
71+
# Rename bounds variables
72+
for bnds_var, out_bnds_name in bounds_rename_map.items():
73+
if bnds_var in self.ds:
74+
self.ds = self.ds.rename({bnds_var: out_bnds_name})
75+
elif bnds_var in self.ds.coords:
76+
self.ds = self.ds.rename({bnds_var: out_bnds_name})
77+
78+
# Update "bounds" attribute in all variables and coordinates
79+
for var in list(self.ds.variables) + list(self.ds.coords):
80+
bounds_attr = self.ds[var].attrs.get("bounds")
81+
if bounds_attr and bounds_attr in bounds_rename_map:
82+
self.ds[var].attrs["bounds"] = bounds_rename_map[bounds_attr]
83+
84+
# Transpose the data variable according to the CMOR dimensions
85+
cmor_dims = self.vocab.variable["dimensions"].split()
86+
transpose_order = [
87+
self.vocab.axes[dim]["out_name"]
88+
for dim in cmor_dims
89+
if "value" not in self.vocab.axes[dim]
90+
]
91+
# Squeeze singleton dimensions if they are not in the transpose order
92+
for dim in self.ds[self.cmor_name].dims:
93+
if dim not in transpose_order and self.ds[self.cmor_name][dim].size == 1:
94+
self.ds[self.cmor_name] = self.ds[self.cmor_name].squeeze(dim)
95+
96+
self.ds[self.cmor_name] = self.ds[self.cmor_name].transpose(*transpose_order)
97+
98+
def update_attributes(self):
99+
self.ds.attrs = {
100+
k: v
101+
for k, v in self.vocab.get_required_global_attributes().items()
102+
if v not in (None, "")
103+
}
104+
105+
required_coords = {
106+
v["out_name"] for v in self.vocab.axes.values() if "value" in v
107+
}.union({v["out_name"] for v in self.vocab.axes.values()})
108+
self.ds = self.ds.drop_vars(
109+
[c for c in self.ds.coords if c not in required_coords], errors="ignore"
110+
)
111+
112+
cmor_attrs = self.vocab.variable
113+
self._check_units(self.cmor_name, cmor_attrs.get("units"))
114+
115+
self.ds[self.cmor_name].attrs.update(
116+
{k: v for k, v in cmor_attrs.items() if v not in (None, "")}
117+
)
118+
var_type = cmor_attrs.get("type", "double")
119+
self.ds[self.cmor_name] = self.ds[self.cmor_name].astype(
120+
self.type_mapping.get(var_type, np.float64)
121+
)
122+
123+
try:
124+
if cmor_attrs.get("valid_min") not in (None, "") and cmor_attrs.get(
125+
"valid_max"
126+
) not in (None, ""):
127+
vmin = self.type_mapping.get(var_type, np.float64)(
128+
cmor_attrs["valid_min"]
129+
)
130+
vmax = self.type_mapping.get(var_type, np.float64)(
131+
cmor_attrs["valid_max"]
132+
)
133+
self._check_range(self.cmor_name, vmin, vmax)
134+
except ValueError as e:
135+
raise ValueError(
136+
f"Failed to validate value range for {self.cmor_name}: {e}"
137+
)
138+
139+
for dim, meta in self.vocab.axes.items():
140+
name = meta["out_name"]
141+
dtype = self.type_mapping.get(meta.get("type", "double"), np.float64)
142+
if name in self.ds:
143+
self._check_units(name, meta.get("units", ""))
144+
if meta.get("standard_name") == "time":
145+
self._check_calendar(name)
146+
original_units = self.ds[name].attrs.get("units", "")
147+
coord_attrs = {
148+
k: v
149+
for k, v in {
150+
"standard_name": meta.get("standard_name"),
151+
"long_name": meta.get("long_name"),
152+
"units": meta.get("units"),
153+
"axis": meta.get("axis"),
154+
"positive": meta.get("positive"),
155+
"valid_min": dtype(meta["valid_min"])
156+
if "valid_min" in meta
157+
else None,
158+
"valid_max": dtype(meta["valid_max"])
159+
if "valid_max" in meta
160+
else None,
161+
}.items()
162+
if v is not None
163+
}
164+
if coord_attrs.get(
165+
"units"
166+
) == "days since ?" and original_units.lower().startswith("days since"):
167+
coord_attrs["units"] = original_units
168+
updated = self.ds[name].astype(dtype)
169+
updated.attrs.update(coord_attrs)
170+
self.ds[name] = updated
171+
elif "value" in meta:
172+
val = meta["value"]
173+
# Handle character type (e.g., string coordinate)
174+
if meta["type"] == "character":
175+
arr = xr.DataArray(
176+
np.array(
177+
val, dtype="S"
178+
), # ensure type is character (byte string)
179+
dims=(),
180+
attrs={
181+
k: v
182+
for k, v in {
183+
"standard_name": meta.get("standard_name"),
184+
"long_name": meta.get("long_name"),
185+
"units": meta.get("units"),
186+
"axis": meta.get("axis"),
187+
"positive": meta.get("positive"),
188+
"valid_min": meta.get("valid_min"),
189+
"valid_max": meta.get("valid_max"),
190+
}.items()
191+
if v is not None
192+
},
193+
)
194+
else:
195+
arr = xr.DataArray(
196+
dtype(val),
197+
dims=(),
198+
attrs={
199+
k: v
200+
for k, v in {
201+
"standard_name": meta.get("standard_name"),
202+
"long_name": meta.get("long_name"),
203+
"units": meta.get("units"),
204+
"axis": meta.get("axis"),
205+
"positive": meta.get("positive"),
206+
"valid_min": dtype(meta["valid_min"])
207+
if "valid_min" in meta
208+
else None,
209+
"valid_max": dtype(meta["valid_max"])
210+
if "valid_max" in meta
211+
else None,
212+
}.items()
213+
if v is not None
214+
},
215+
)
216+
self.ds = self.ds.assign_coords({name: arr})

0 commit comments

Comments
 (0)