Merge branch 'develop' of github.com:ecmwf/anemoi-datasets into develop

Author: b8raoult

.github/CODEOWNERS

@@ -1,6 +1,6 @@
 # CODEOWNERS file
 
 # Protect workflow files
-/.github/ @theissenhelen @jesperdramsch @gmertes @b8raoult @floriankrb
-/.pre-commit-config.yaml @theissenhelen @jesperdramsch @gmertes @b8raoult @floriankrb
-/pyproject.toml @theissenhelen @jesperdramsch @gmertes @b8raoult @floriankrb
+/.github/ @theissenhelen @jesperdramsch @gmertes @b8raoult @floriankrb @anaprietonem @HCookie @JPXKQX @mchantry
+/.pre-commit-config.yaml @theissenhelen @jesperdramsch @gmertes @b8raoult @floriankrb @anaprietonem @HCookie @JPXKQX @mchantry
+/pyproject.toml @theissenhelen @jesperdramsch @gmertes @b8raoult @floriankrb @anaprietonem @HCookie @JPXKQX @mchantry

.github/ISSUE_TEMPLATE/bug_report.md

@@ -9,12 +9,15 @@ assignees: ''
 **Describe the bug**
 A clear and concise description of what the bug is.
 
+** Version number **
+I am using the following versions/branch/sha1 of the anemoi packages
+(alternatively the output of `pip freeze`)
+
 **To Reproduce**
 Steps to reproduce the behavior:
 1. Go to '...'
-2. Click on '....'
-3. Scroll down to '....'
-4. See error
+2. Run this '....'
+3. See error
 
 **URL to sample input data**
 Provide a URL to a sample input data, or attach a file to that report if it is small enough.
@@ -25,16 +28,5 @@ A clear and concise description of what you expected to happen.
 **Screenshots**
 If applicable, add screenshots to help explain your problem.
 
-**Desktop (please complete the following information):**
- - OS: [e.g. iOS]
- - Browser [e.g. chrome, safari]
- - Version [e.g. 22]
-
-**Smartphone (please complete the following information):**
- - Device: [e.g. iPhone6]
- - OS: [e.g. iOS8.1]
- - Browser [e.g. stock browser, safari]
- - Version [e.g. 22]
-
 **Additional context**
 Add any other context about the problem here.

.github/ci-config.yml

@@ -1,9 +1,3 @@
-dependencies: |
-  ecmwf/ecbuild
-  MathisRosenhauer/libaec@master
-  ecmwf/eccodes
-  ecmwf/eckit
-  ecmwf/odc
 dependency_branch: develop
 parallelism_factor: 8
 self_build: false # Only for python packages

.github/ci-hpc-config.yml

@@ -2,17 +2,8 @@ build:
   python: '3.10'
   modules:
     - ninja
-  dependencies:
-    - ecmwf/ecbuild@develop
-    - ecmwf/eccodes@develop
-    - ecmwf/eckit@develop
-    - ecmwf/odc@develop
   python_dependencies:
     - ecmwf/anemoi-utils@develop
-    - ecmwf/earthkit-data@develop
-    - ecmwf/earthkit-meteo@develop
-    - ecmwf/earthkit-geo@develop
   parallel: 64
-
   pytest_cmd: |
     python -m pytest -vv -m 'not notebook and not no_cache_init' --cov=. --cov-report=xml

CHANGELOG.md

@@ -10,13 +10,31 @@ Keep it human-readable, your future self will thank you!
 
 ## [Unreleased](https://github.com/ecmwf/anemoi-datasets/compare/0.5.7...HEAD)
 
+
+### Added
+
+- Add anemoi-transform link to documentation
+- Various bug fixes
+- Control compatibility check in xy/zip
+- Add `merge` feature
+
+### Changed
+
+- Remove upstream dependencies from downstream-ci workflow (temporary) (#83)
+
 ## [0.5.7](https://github.com/ecmwf/anemoi-datasets/compare/0.5.6...0.5.7) - 2024-10-09
 
+### Changed
+
+- Add support to fill missing dates
+
 ## [Allow for unknown CF coordinates](https://github.com/ecmwf/anemoi-datasets/compare/0.5.5...0.5.6) - 2024-10-04
 
-- Update documentation
+### Changed
 
+- Add `variables_metadata` entry in the dataset metadata
 - Update documentation
+
 ### Changed
 
 - Add `variables_metadata` entry in the dataset metadata

docs/conf.py

@@ -107,6 +107,10 @@
         "https://anemoi-registry.readthedocs.io/en/latest/",
         ("../../anemoi-registry/docs/_build/html/objects.inv", None),
     ),
+    "anemoi-transform": (
+        "https://anemoi-transform.readthedocs.io/en/latest/",
+        ("../../anemoi-transform/docs/_build/html/objects.inv", None),
+    ),
 }
 
 

docs/index.rst

@@ -121,6 +121,7 @@ datasets <building-introduction>`.
 *****************
 
 -  :ref:`anemoi-utils <anemoi-utils:index-page>`
+-  :ref:`anemoi-transform <anemoi-transform:index-page>`
 -  :ref:`anemoi-datasets <anemoi-datasets:index-page>`
 -  :ref:`anemoi-models <anemoi-models:index-page>`
 -  :ref:`anemoi-graphs <anemoi-graphs:index-page>`

docs/using/code/fill_missing_dates1_.py

@@ -0,0 +1 @@
+ds = open_dataset(dataset, fill_missing_dates="interpolate")

docs/using/code/fill_missing_dates2_.py

@@ -0,0 +1 @@
+ds = open_dataset(dataset, fill_missing_dates="closest")

docs/using/code/missing_dates_.py

@@ -0,0 +1 @@
+ds = open_dataset(dataset, set_missing_dates=["2010-01-01T12:00:00", "2010-02-01T12:00:00"])

docs/using/code/set_missing_dates_.py

@@ -4,6 +4,25 @@
  Managing missing dates
 ########################
 
+**************************************************
+ Filling the missing dates with artificial values
+**************************************************
+
+When you have missing dates in a dataset, you can fill them with
+artificial values. You can either fill them with values that are the
+result of a linear interpolation between the two closest dates:
+
+.. literalinclude:: code/fill_missing_dates1_.py
+
+Or you can select the copy the value of the closest date:
+
+.. literalinclude:: code/fill_missing_dates2_.py
+
+if the missing date is exactly in the middle of two dates, the library
+will choose that value of the largest date. You can change this behavior
+by setting the ``closest`` parameter to ``'down'`` or ``'up'``
+explicitly.
+
 ************************************************
  Skipping missing when iterating over a dataset
 ************************************************
@@ -72,7 +91,7 @@ the datasets to make the dates contiguous.
  Debugging
 ***********
 
-You can set missing dates using the ``missing_dates`` option. This
+You can set missing dates using the ``set_missing_dates`` option. This
 option is for debugging purposes only.
 
-.. literalinclude:: code/missing_dates_.py
+.. literalinclude:: code/set_missing_dates_.py

docs/using/missing.rst

@@ -50,7 +50,7 @@ dynamic = [
   "version",
 ]
 dependencies = [
-  "anemoi-utils[provenance]>=0.3.15",
+  "anemoi-utils[provenance]>=0.3.18",
   "cfunits",
   "numpy",
   "pyyaml",
@@ -60,43 +60,20 @@ dependencies = [
 ]
 
 optional-dependencies.all = [
-  "boto3",
-  "earthkit-data[mars]>=0.9",
-  "earthkit-geo>=0.2",
-  "earthkit-meteo",
-  "ecmwflibs>=0.6.3",
-  "entrypoints",
-  "gcsfs",
-  "kerchunk",
-  "pyproj",
-  "requests",
+  "anemoi-datasets[create,remote,xarray]",
 ]
 
 optional-dependencies.create = [
-  "earthkit-data[mars]>=0.9",
+  "earthkit-data[mars]>=0.10.7",
   "earthkit-geo>=0.2",
   "earthkit-meteo",
-  "ecmwflibs>=0.6.3",
+  "eccodes>=2.38.1",
   "entrypoints",
   "pyproj",
 ]
 
 optional-dependencies.dev = [
-  "boto3",
-  "earthkit-data[mars]>=0.9",
-  "earthkit-geo>=0.2",
-  "earthkit-meteo",
-  "ecmwflibs>=0.6.3",
-  "entrypoints",
-  "gcsfs",
-  "kerchunk",
-  "nbsphinx",
-  "pandoc",
-  "pyproj",
-  "pytest",
-  "requests",
-  "sphinx",
-  "sphinx-rtd-theme",
+  "anemoi-datasets[all,docs,tests]",
 ]
 
 optional-dependencies.docs = [

pyproject.toml

@@ -56,11 +56,14 @@ def __init__(self, field, format):
         self.format = format
         self.bits = re.findall(r"{(\w+)}", format)
 
-    def metadata(self, key, **kwargs):
-        value = self.field.metadata(key, **kwargs)
-        if "{" + key + "}" in self.format:
-            bits = {b: self.field.metadata(b, **kwargs) for b in self.bits}
-            return self.format.format(**bits)
+    def metadata(self, *args, **kwargs):
+        value = self.field.metadata(*args, **kwargs)
+        if args:
+            assert len(args) == 1
+            key = args[0]
+            if "{" + key + "}" in self.format:
+                bits = {b: self.field.metadata(b, **kwargs) for b in self.bits}
+                return self.format.format(**bits)
         return value
 
     def __getattr__(self, name):

src/anemoi/datasets/create/functions/filters/rename.py

@@ -370,12 +370,15 @@ def accumulations(context, dates, **request):
 
     user_accumulation_period = request.pop("accumulation_period", 6)
 
+    # If `data_accumulation_period` is not set, this means that the accumulations are from the start
+    # of the forecast.
+
     KWARGS = {
         ("od", "oper"): dict(patch=_scda),
         ("od", "elda"): dict(base_times=(6, 18)),
         ("ea", "oper"): dict(data_accumulation_period=1, base_times=(6, 18)),
         ("ea", "enda"): dict(data_accumulation_period=3, base_times=(6, 18)),
-        ("rr", "oper"): dict(data_accumulation_period=3, base_times=(0, 3, 6, 9, 12, 15, 18, 21)),
+        ("rr", "oper"): dict(base_times=(0, 3, 6, 9, 12, 15, 18, 21)),
         ("l5", "oper"): dict(data_accumulation_period=1, base_times=(0,)),
     }
 

src/anemoi/datasets/create/functions/sources/accumulations.py

@@ -6,23 +6,9 @@
 # granted to it by virtue of its status as an intergovernmental organisation
 # nor does it submit to any jurisdiction.
 #
-import datetime
-import itertools
+
 import logging
-import math
-import time
-from collections import defaultdict
 from copy import deepcopy
-from functools import cached_property
-from functools import wraps
-
-import numpy as np
-from anemoi.utils.dates import as_datetime as as_datetime
-from anemoi.utils.dates import frequency_to_timedelta as frequency_to_timedelta
-
-from anemoi.datasets.dates import DatesProvider as DatesProvider
-from anemoi.datasets.fields import FieldArray as FieldArray
-from anemoi.datasets.fields import NewValidDateTimeField as NewValidDateTimeField
 
 from .trace import trace_select
 

src/anemoi/datasets/create/input/__init__.py

@@ -33,9 +33,38 @@
 def _fields_metatata(variables, cube):
     assert isinstance(variables, tuple), variables
 
+    def _merge(md1, md2):
+        assert set(md1.keys()) == set(md2.keys()), (set(md1.keys()), set(md2.keys()))
+        result = {}
+        for k in md1.keys():
+            v1 = md1[k]
+            v2 = md2[k]
+
+            if v1 == v2:
+                result[k] = v1
+                continue
+
+            if isinstance(v1, list):
+                assert v2 not in v1, (v1, v2)
+                result[k] = sorted(v1 + [v2])
+                continue
+
+            if isinstance(v2, list):
+                assert v1 not in v2, (v1, v2)
+                result[k] = sorted(v2 + [v1])
+                continue
+
+            result[k] = sorted([v1, v2])
+
+        return result
+
     result = {}
-    for i, c in enumerate(cube.iterate_cubelets()):
-        assert c._coords_names[1] == variables[i], (c._coords_names[1], variables[i])
+    i = -1
+    for c in cube.iterate_cubelets():
+
+        if i == -1 or c._coords_names[1] != variables[i]:
+            i += 1
+
         f = cube[c.coords]
         md = f.metadata(namespace="mars")
         if not md:
@@ -49,7 +78,10 @@ def _fields_metatata(variables, cube):
             md["param"] = str(f.metadata("paramId", default="unknown"))
             # assert md['param'] != 'unknown', (md, f.metadata('param'))
 
-        result[variables[i]] = md
+        if variables[i] in result:
+            result[variables[i]] = _merge(md, result[variables[i]])
+        else:
+            result[variables[i]] = md
 
     assert i + 1 == len(variables), (i + 1, len(variables))
     return result

src/anemoi/datasets/create/input/result.py

@@ -59,6 +59,7 @@ def select(self, group_of_dates):
         )
 
     def __repr__(self):
+        # raise NotImplementedError(f"Not implemented in {self.__class__.__name__}")
         return super().__repr__(self.previous_step, _inline_=str(self.kwargs))
 
 

src/anemoi/datasets/create/input/step.py

@@ -148,6 +148,7 @@ def concat_factory(args, kwargs):
 
     datasets = kwargs.pop("concat")
     fill_missing_gaps = kwargs.pop("fill_missing_gaps", False)
+
     assert isinstance(datasets, (list, tuple))
     assert len(args) == 0