Add CMS Global Mangrove Canopy dataset

docs/api/datasets.rst

@@ -32,6 +32,11 @@ Chesapeake Bay High-Resolution Land Cover Project
 .. autoclass:: ChesapeakeWV
 .. autoclass:: ChesapeakeCVPR
 
+CMS Global Mangrove Canopy Dataset
+^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
+
+.. autoclass:: CMS_Global_Mangrove_Canopy
+
 Cropland Data Layer (CDL)
 ^^^^^^^^^^^^^^^^^^^^^^^^^
 

tests/data/cms_mangrove_canopy/data.py

@@ -0,0 +1,66 @@
+#!/usr/bin/env python3
+
+# Copyright (c) Microsoft Corporation. All rights reserved.
+# Licensed under the MIT License.
+
+import hashlib
+import os
+import random
+import shutil
+
+import numpy as np
+import rasterio
+
+np.random.seed(0)
+random.seed(0)
+
+SIZE = 64
+
+
+files = [
+    {"image": "Mangrove_agb_Angola.tif"},
+    {"image": "Mangrove_hba95_Angola.tif"},
+    {"image": "Mangrove_hmax95_Angola.tif"},
+]
+
+
+def create_file(path: str, dtype: str, num_channels: int) -> None:
+    profile = {}
+    profile["driver"] = "GTiff"
+    profile["dtype"] = dtype
+    profile["count"] = num_channels
+    profile["crs"] = "epsg:4326"
+    profile["transform"] = rasterio.transform.from_bounds(0, 0, 1, 1, 1, 1)
+    profile["height"] = SIZE
+    profile["width"] = SIZE
+
+    Z = np.random.randint(
+        np.iinfo(profile["dtype"]).max, size=(1, SIZE, SIZE), dtype=profile["dtype"]
+    )
+    src = rasterio.open(path, "w", **profile)
+    src.write(Z)
+
+
+if __name__ == "__main__":
+    directory = "CMS_Global_Map_Mangrove_Canopy_1665"
+
+    # Remove old data
+    if os.path.isdir(directory):
+        shutil.rmtree(directory)
+
+    os.makedirs(os.path.join(directory, "data"), exist_ok=True)
+
+    for file_dict in files:
+        # Create mask file
+        path = file_dict["image"]
+        create_file(
+            os.path.join(directory, "data", path), dtype="int32", num_channels=1
+        )
+
+    # Compress data
+    shutil.make_archive(directory.replace(".zip", ""), "zip", ".", directory)
+
+    # Compute checksums
+    with open(directory + ".zip", "rb") as f:
+        md5 = hashlib.md5(f.read()).hexdigest()
+        print(f"{directory}: {md5}")

tests/datasets/test_cms_mangrove_canopy.py

@@ -0,0 +1,89 @@
+# Copyright (c) Microsoft Corporation. All rights reserved.
+# Licensed under the MIT License.
+
+import os
+import shutil
+from pathlib import Path
+from typing import Generator
+
+import pytest
+import torch
+import torch.nn as nn
+from _pytest.monkeypatch import MonkeyPatch
+from rasterio.crs import CRS
+
+from torchgeo.datasets import (
+    CMS_Global_Mangrove_Canopy,
+    IntersectionDataset,
+    UnionDataset,
+)
+
+
+def download_url(url: str, root: str, *args: str, **kwargs: str) -> None:
+    shutil.copy(url, root)
+
+
+class TestCMS_Global_Mangrove_Canopy:
+    @pytest.fixture
+    def dataset(
+        self, monkeypatch: Generator[MonkeyPatch, None, None], tmp_path: Path
+    ) -> CMS_Global_Mangrove_Canopy:
+        zipfile = "CMS_Global_Map_Mangrove_Canopy_1665.zip"
+        monkeypatch.setattr(  # type: ignore[attr-defined]
+            CMS_Global_Mangrove_Canopy, "zipfile", zipfile
+        )
+
+        md5 = "45d894d4516d91212b70fd9a803e28cd"
+        monkeypatch.setattr(  # type: ignore[attr-defined]
+            CMS_Global_Mangrove_Canopy, "md5", md5
+        )
+
+        root = os.path.join("tests", "data", "cms_mangrove_canopy")
+        transforms = nn.Identity()  # type: ignore[attr-defined]
+        country = "Angola"
+
+        return CMS_Global_Mangrove_Canopy(
+            root, country=country, transforms=transforms, checksum=True
+        )
+
+    def test_getitem(self, dataset: CMS_Global_Mangrove_Canopy) -> None:
+        x = dataset[dataset.bounds]
+        assert isinstance(x, dict)
+        assert isinstance(x["crs"], CRS)
+        assert isinstance(x["mask"], torch.Tensor)
+
+    def test_no_dataset(self) -> None:
+        with pytest.raises(RuntimeError, match="Dataset not found in."):
+            CMS_Global_Mangrove_Canopy(root="/test")
+
+    def test_already_downloaded(self, tmp_path: Path) -> None:
+        pathname = os.path.join(
+            "tests",
+            "data",
+            "cms_mangrove_canopy",
+            "CMS_Global_Map_Mangrove_Canopy_1665.zip",
+        )
+        root = str(tmp_path)
+        shutil.copy(pathname, root)
+        CMS_Global_Mangrove_Canopy(root, country="Angola")
+
+    def test_invalid_country(self) -> None:
+        with pytest.raises(AssertionError):
+            CMS_Global_Mangrove_Canopy(country="fakeCountry")
+
+    def test_invalid_measurement(self) -> None:
+        with pytest.raises(AssertionError):
+            CMS_Global_Mangrove_Canopy(measurement="wrongMeasurement")
+
+    def test_and(self, dataset: CMS_Global_Mangrove_Canopy) -> None:
+        ds = dataset & dataset
+        assert isinstance(ds, IntersectionDataset)
+
+    def test_or(self, dataset: CMS_Global_Mangrove_Canopy) -> None:
+        ds = dataset | dataset
+        assert isinstance(ds, UnionDataset)
+
+    def test_plot(self, dataset: CMS_Global_Mangrove_Canopy) -> None:
+        query = dataset.bounds
+        x = dataset[query]
+        dataset.plot(x["mask"])

torchgeo/datasets/__init__.py

@@ -21,6 +21,7 @@
     ChesapeakeVA,
     ChesapeakeWV,
 )
+from .cms_mangrove_canopy import CMS_Global_Mangrove_Canopy
 from .cowc import COWC, COWCCounting, COWCDetection
 from .cv4a_kenya_crop_type import CV4AKenyaCropType
 from .cyclone import TropicalCycloneWindEstimation
@@ -96,6 +97,7 @@
     "ChesapeakeVA",
     "ChesapeakeWV",
     "ChesapeakeCVPR",
+    "CMS_Global_Mangrove_Canopy",
     "Landsat",
     "Landsat1",
     "Landsat2",