Fix outstanding issues

.github/workflows/unit-tests.yml

@@ -0,0 +1,44 @@
+name: Unit Tests
+
+on:
+  push:
+  schedule:
+    - cron: "0 2 * * 0"
+  workflow_dispatch:
+
+permissions:
+  contents: read
+
+jobs:
+  unit-tests:
+    runs-on: ubuntu-latest
+    timeout-minutes: 30
+
+    steps:
+      - name: Checkout
+        uses: actions/checkout@v4
+
+      - name: Set up Python
+        uses: actions/setup-python@v5
+        with:
+          python-version: "3.10"
+          cache: pip
+          cache-dependency-path: |
+            requirements/dev-requirements.txt
+
+      - name: Upgrade pip
+        run: python -m pip install --upgrade pip
+
+      - name: Install development dependencies
+        run: python -m pip install -r requirements/dev-requirements.txt
+
+      - name: Install unit-test runtime dependencies
+        run: |
+          python -m pip install "numpy<2" xarray matplotlib cmocean gsw
+          python -m pip install "https://github.com/sdat2/pyxpcm/archive/sof-agu.zip"
+          python -m pip install --upgrade "numpy<2"
+
+      - name: Run unit tests
+        env:
+          MPLBACKEND: Agg
+        run: python -m unittest discover -s src/tests -p "test_*.py" -v

README.md

@@ -62,11 +62,23 @@ of individual eddy-like features (such as the Agulhas rings).
     make env
     ```
 
+    or with `micromamba`
+
+    ```bash
+    micromamba create -f requirements/environment.yml -p ./env
+    ```
+
 - Activate the environment in conda:
 
-     ```bash
-     conda activate ./env
-     ```
+    ```bash
+    conda activate ./env
+
+    micromamba activate ./env
+    ```
+
+    ```bash
+    micromamba activate ./env 
+    ```
 
 - Change the settings in `src.constants` to set download location etc.
 

requirements/dev-requirements.txt

@@ -2,7 +2,7 @@
 # As a standard they include certain formatters and linters.
 
 # local package
--e ../.
+-e .
 
 # external requirements (mostly linters and formatters)
 flake8              # flake8 linter

requirements/environment.yml

@@ -6,8 +6,8 @@ channels:
 dependencies:
   - python=3.8
   - xarray==0.15
-  - cartopy
-  - matplotlib==3.2.2
+  - cartopy==0.18
+  - matplotlib
   - pip
   - pip:
     - -r dev-requirements.txt

requirements/requirements.txt

@@ -10,7 +10,7 @@ wandb
 xarray==0.15
 netcdf4
 scikit-learn
-matplotlib==3.2.2
+matplotlib
 
 # pynio
 numba

src/animate.py

@@ -1,4 +1,4 @@
-"""Animate da."""
+"""Animate dataarray."""
 import os
 from typing import Callable
 import numpy as np

src/constants.py

@@ -2,51 +2,67 @@
 
 # Place all your constants here
 import os
+from typing import Literal, List, Dict
 import numpy as np
-import pathlib
 from sys import platform
 import cmocean.cm as cmo
 
 # Note: constants should be UPPER_CASE
 
-# Basic location defaults, to be referenced from here:
-constants_path = os.path.realpath(__file__)
-SRC_PATH = os.path.dirname(constants_path)
-PROJECT_PATH = os.path.dirname(SRC_PATH)
-DATA_PATH = os.path.join(PROJECT_PATH, "nc")
-FIGURE_PATH = os.path.join(PROJECT_PATH, "figures")
-KO_PATH = os.path.join(SRC_PATH, "data", "kim_(&orsi)_altimetric_fronts")
+# TODO move some of these to a config file.
 
-# Data directory on GWS
-GWS_DATA_DIR = pathlib.Path("/gws/nopw/j04/ai4er/users/sdat2/OLD")
+# Basic location defaults, to be referenced from here:
+constants_path: str = os.path.realpath(__file__)
+SRC_PATH: str = os.path.dirname(constants_path)
+PROJECT_PATH: str = os.path.dirname(SRC_PATH)
+DATA_PATH: str = os.path.join(PROJECT_PATH, "nc")
+FIGURE_PATH: str = os.path.join(PROJECT_PATH, "figures")
+KO_PATH: str = os.path.join(SRC_PATH, "data", "kim_(&orsi)_altimetric_fronts")
 
 # Figure type
-FIGURE_TYPE = ".png"
+FIGURE_TYPE: Literal[".png", ".pdf"] = ".png"
 
 # start ****DATA LOCATION section***
-# This will certainly need to be changed on your macine
+# This will certainly need to be changed on your machine
+GEN_ROOT: str = DATA_PATH
+DEFAULT_NC: str = os.path.join(GEN_ROOT, "i-metric-joint-k-5-d-3.nc")
 
-# Paths to BSOSE (unique to Jasmin)
+# Keep the historical Linux vs Darwin defaults, but prefer folders that
+# already exist so the code remains robust when moved between machines.
 if platform in ["Linux", "linux"]:
-    GEN_DATA_PATH: str = os.path.join(GWS_DATA_DIR, "bsose_data")
-    BSOSE_PATH: str = os.path.join(GEN_DATA_PATH, "bsose_stuv")
-    DEFAULT_NC: str = (
-        str(GWS_DATA_DIR) + "/nc/i-metric-joint-k-5-d-3.nc"  # not valid in jasmin.
-    )
-
-# Paths to different BSOSE-i106 files (unique to my machine):
-elif platform in ["Darwin", "darwin"]:
-    BSOSE_PATH: str = os.path.join("/Users", "simon", "bsose_monthly")
-    GEN_DATA_PATH: str = BSOSE_PATH
-    DEFAULT_NC: str = (
-        "~/pyxpcm_sithom/nc/i-metric-joint-k-5-d-3.nc"  # not valid in jasmin.
-    )
-
+    _preferred_data_dirs = [
+        os.path.join(GEN_ROOT, "bsose_data"),
+        os.path.join(GEN_ROOT, "bsose_monthly"),
+    ]
 else:
-    assert False
+    _preferred_data_dirs = [
+        os.path.join(GEN_ROOT, "bsose_monthly"),
+        os.path.join(GEN_ROOT, "bsose_data"),
+    ]
+
+GEN_DATA_PATH: str = next(
+    (path for path in _preferred_data_dirs if os.path.isdir(path)),
+    _preferred_data_dirs[0],
+)
+
+_bsose_candidates = [
+    os.path.join(GEN_DATA_PATH, "bsose_stuv"),
+    GEN_DATA_PATH,
+]
+BSOSE_PATH: str = next(
+    (path for path in _bsose_candidates if os.path.isdir(path)),
+    _bsose_candidates[0],
+)
 
 # end ****DATA LOCATION section***
 
+
+os.makedirs(DATA_PATH, exist_ok=True)
+os.makedirs(FIGURE_PATH, exist_ok=True)
+os.makedirs(KO_PATH, exist_ok=True)
+os.makedirs(GEN_DATA_PATH, exist_ok=True)
+os.makedirs(BSOSE_PATH, exist_ok=True)
+
 # Salt, Theta, Uvel, Vvel
 SALT_FILE: str = os.path.join(BSOSE_PATH, "bsose_i106_2008to2012_monthly_Salt.nc")
 THETA_FILE: str = os.path.join(BSOSE_PATH, "bsose_i106_2008to2012_monthly_Theta.nc")
@@ -66,9 +82,9 @@
 
 # Particular names within BSOSE-i106
 DEPTH_NAME: str = D_COORD
-USELESS_LIST: list = ["iter", "Depth", "rA", "drF", "hFacC"] # list of variables from BSOSE to discard before processing
-VAR_NAME_LIST: list = ["SALT", "THETA"] # variables used in to fit the pcm model on
-FEATURES_D: dict = {"THETA": "THETA", "SALT": "SALT"} # Mapping for within pyxpcm
+USELESS_LIST: List[str] = ["iter", "Depth", "rA", "drF", "hFacC"] # list of variables from BSOSE to discard before processing
+VAR_NAME_LIST: List[str] = ["SALT", "THETA"] # variables used in to fit the pcm model on
+FEATURES_D: Dict[str, str] = {"THETA": "THETA", "SALT": "SALT"} # Mapping for within pyxpcm
 
 # Naming of intermediate files
 INTERP_FILE_NAME: str = os.path.join(DATA_PATH, "interp.nc")
@@ -82,7 +98,7 @@
 # random variables used locally to make it reproducible.
 MIN_DEPTH: float = 300  # the depth of the minimum cut off (m)
 MAX_DEPTH: float = 2000  # the depth of the maximum cut off (m)
-K_LIST: list = [5, 4, 2, 10]  # K's to make when running batch script.
+K_LIST: List[int] = [5, 4, 2, 10]  # K's to make when running batch script.
 K_CLUSTERS: int = 5  # number of clusters for the main example figure.
 D_PCS: int = 3  # number of principal components to be used.
 EXAMPLE_TIME_INDEX: int = 40  # the default time to go for.
@@ -96,9 +112,10 @@
 CLUST_COLORS: str = "Set1" # "Dark1"
 
 # Move plots to location
-FINAL_LOC: str = "../FBSO/images"
 
+FINAL_LOC: str = os.path.join(PROJECT_PATH, "images") # "../FBSO/images"
+os.makedirs(FINAL_LOC, exist_ok=True)
 
-# infor for profile plots
-ZS = [-x for x in range(300, 2000, 10)] # Z levels.
-LZ = len(ZS) # number of Z levels.
+# info for profile plots
+ZS: List[int] = [-x for x in range(300, 2000, 10)] # Z levels.
+LZ: int = len(ZS) # number of Z levels.

src/data_loading/io_names.py

@@ -21,7 +21,7 @@ def return_pair_i_metric(
         pca (int, optional): Number of principal components. Defaults to cst.D_PCS.
         save_nc (bool, optional): Whether or not to save the resulting dataset.
             Defaults to True.
-        t_index (int, optional): time index cst.EXAMPLE_TIME_INDEX.
+        t_index (int, optional): Time index cst.EXAMPLE_TIME_INDEX.
 
     Returns:
         xr.DataArray: pair i metric.
@@ -57,12 +57,9 @@ def return_name(k_clusters: int, pca_components: int) -> str:
     Returns:
         str: file names.
     """
-    return (
-        str(cst.GWS_DATA_DIR)
-        + "/nc/i-metric-joint-k-"
-        + str(k_clusters)
-        + "-d-"
-        + str(pca_components)
+    return os.path.join(
+        cst.DATA_PATH,
+        "i-metric-joint-k-" + str(k_clusters) + "-d-" + str(pca_components),
     )
 
 
@@ -76,17 +73,13 @@ def return_plot_folder(k_clusters: int, pca_components: int) -> str:
     Returns:
         str: file names.
     """
-    folder = (
-        "../FBSO-Report/"
-        + "images/i-metric-joint-k-"
-        + str(k_clusters)
-        + "-d-"
-        + str(pca_components)
-        + "/"
+    folder = os.path.join(
+        cst.FIGURE_PATH,
+        "i-metric-joint-k-" + str(k_clusters) + "-d-" + str(pca_components),
     )
     if not os.path.exists(folder):
         os.makedirs(folder)
-    return folder
+    return os.path.join(folder, "")
 
 
 def return_folder(k_clusters: int, pca_components: int) -> str:
@@ -100,7 +93,7 @@ def return_folder(k_clusters: int, pca_components: int) -> str:
         str: file names.
 
     """
-    folder = return_name(k_clusters, pca_components) + "/"
+    folder = os.path.join(return_name(k_clusters, pca_components), "")
     if not os.path.exists(folder):
         os.makedirs(folder)
     return folder
@@ -117,12 +110,9 @@ def _return_pair_name(k_clusters: int, pca_components: int) -> str:
         str: file names.
 
     """
-    return (
-        str(cst.GWS_DATA_DIR)
-        + "nc/pair-i-metric-k-"
-        + str(k_clusters)
-        + "-d-"
-        + str(pca_components)
+    return os.path.join(
+        cst.DATA_PATH,
+        "pair-i-metric-k-" + str(k_clusters) + "-d-" + str(pca_components),
     )
 
 
@@ -137,14 +127,7 @@ def _return_pair_folder(k_clusters: int, pca_components: int) -> str:
         str: file names.
 
     """
-    folder = (
-        str(cst.GWS_DATA_DIR)
-        + "/nc/pair-i-metric-k-"
-        + str(k_clusters)
-        + "-d-"
-        + str(pca_components)
-        + "/"
-    )
+    folder = os.path.join(_return_pair_name(k_clusters, pca_components), "")
     if not os.path.exists(folder):
         os.makedirs(folder)
     return folder

src/data_loading/xr_loader.py

@@ -42,7 +42,6 @@ def _old_order_indexes(dataarray: xr.DataArray, index_list: list) -> np.ndarray:
 
     Returns:
         np.ndarray: dataarray_values.
-
     """
 
     coords_list = []

src/make_figures.py

@@ -1,4 +1,5 @@
-"""Make figures: run through all the paper figures and make them.
+"""
+Make figures: run through all the paper figures and make them.
 
 Takes roughly 5 minutes the first time it is run.
 """
@@ -23,7 +24,7 @@
 
 
 @twr.timeit
-def make_all_figures():
+def make_all_figures() -> None:
     """
     Make all the figures in the paper in a sequence.