Promote pandas to a required dependency

pyproject.toml

@@ -51,6 +51,8 @@ dependencies = [
     "matplotlib>=3.3.3",
     # numpy 1.19.3+ required by Python 3.9 wheels
     "numpy>=1.19.3",
+    # pandas 1.1.3+ required for Python 3.9 wheels
+    "pandas>=1.1.3",
     # pillow 8+ required for Python 3.9 wheels
     "pillow>=8",
     # pyproj 3+ required for Python 3.9 wheels
@@ -82,8 +84,6 @@ datasets = [
     "laspy>=2",
     # opencv-python 4.4.0.46+ required for Python 3.9 wheels
     "opencv-python>=4.4.0.46",
-    # pandas 1.1.3+ required for Python 3.9 wheels
-    "pandas>=1.1.3",
     # pycocotools 2.0.5+ required for cython 3+ support
     "pycocotools>=2.0.5",
     # pyvista 0.29+ required for to avoid segfault during testing

requirements/datasets.txt

@@ -2,7 +2,6 @@
 h5py==3.9.0
 laspy==2.5.1
 opencv-python==4.8.0.76
-pandas==2.1.1
 pycocotools==2.0.7
 pyvista==0.42.2
 radiant-mlhub==0.4.1

requirements/min-reqs.old

@@ -9,6 +9,7 @@ lightly==1.4.4
 lightning==1.8.0
 matplotlib==3.3.3
 numpy==1.19.3
+pandas==1.1.3
 pillow==8.0.0
 pyproj==3.0.0
 rasterio==1.2.0
@@ -24,7 +25,6 @@ torchvision==0.13.0
 h5py==3.0.0
 laspy==2.0.0
 opencv-python==4.4.0.46
-pandas==1.1.3
 pycocotools==2.0.5
 pyvista==0.29.0
 radiant-mlhub==0.3.0

requirements/required.txt

@@ -9,6 +9,7 @@ lightly==1.4.19
 lightning==2.0.9
 matplotlib==3.8.0
 numpy==1.26.0
+pandas==2.1.1
 pillow==10.0.1
 pyproj==3.6.1
 rasterio==1.3.8

tests/datasets/test_eddmaps.py

@@ -1,18 +1,13 @@
 # Copyright (c) Microsoft Corporation. All rights reserved.
 # Licensed under the MIT License.
 
-import builtins
 import os
 from pathlib import Path
-from typing import Any
 
 import pytest
-from pytest import MonkeyPatch
 
 from torchgeo.datasets import BoundingBox, EDDMapS, IntersectionDataset, UnionDataset
 
-pytest.importorskip("pandas", minversion="1.1.3")
-
 
 class TestEDDMapS:
     @pytest.fixture(scope="class")
@@ -39,26 +34,6 @@ def test_no_data(self, tmp_path: Path) -> None:
         with pytest.raises(FileNotFoundError, match="Dataset not found"):
             EDDMapS(str(tmp_path))
 
-    @pytest.fixture
-    def mock_missing_module(self, monkeypatch: MonkeyPatch) -> None:
-        import_orig = builtins.__import__
-
-        def mocked_import(name: str, *args: Any, **kwargs: Any) -> Any:
-            if name == "pandas":
-                raise ImportError()
-            return import_orig(name, *args, **kwargs)
-
-        monkeypatch.setattr(builtins, "__import__", mocked_import)
-
-    def test_mock_missing_module(
-        self, dataset: EDDMapS, mock_missing_module: None
-    ) -> None:
-        with pytest.raises(
-            ImportError,
-            match="pandas is not installed and is required to use this dataset",
-        ):
-            EDDMapS(dataset.root)
-
     def test_invalid_query(self, dataset: EDDMapS) -> None:
         query = BoundingBox(0, 0, 0, 0, 0, 0)
         with pytest.raises(

tests/datasets/test_gbif.py

@@ -1,18 +1,13 @@
 # Copyright (c) Microsoft Corporation. All rights reserved.
 # Licensed under the MIT License.
 
-import builtins
 import os
 from pathlib import Path
-from typing import Any
 
 import pytest
-from pytest import MonkeyPatch
 
 from torchgeo.datasets import GBIF, BoundingBox, IntersectionDataset, UnionDataset
 
-pytest.importorskip("pandas", minversion="1.1.3")
-
 
 class TestGBIF:
     @pytest.fixture(scope="class")
@@ -39,26 +34,6 @@ def test_no_data(self, tmp_path: Path) -> None:
         with pytest.raises(FileNotFoundError, match="Dataset not found"):
             GBIF(str(tmp_path))
 
-    @pytest.fixture
-    def mock_missing_module(self, monkeypatch: MonkeyPatch) -> None:
-        import_orig = builtins.__import__
-
-        def mocked_import(name: str, *args: Any, **kwargs: Any) -> Any:
-            if name == "pandas":
-                raise ImportError()
-            return import_orig(name, *args, **kwargs)
-
-        monkeypatch.setattr(builtins, "__import__", mocked_import)
-
-    def test_mock_missing_module(
-        self, dataset: GBIF, mock_missing_module: None
-    ) -> None:
-        with pytest.raises(
-            ImportError,
-            match="pandas is not installed and is required to use this dataset",
-        ):
-            GBIF(dataset.root)
-
     def test_invalid_query(self, dataset: GBIF) -> None:
         query = BoundingBox(0, 0, 0, 0, 0, 0)
         with pytest.raises(

tests/datasets/test_idtrees.py

@@ -18,7 +18,6 @@
 import torchgeo.datasets.utils
 from torchgeo.datasets import IDTReeS
 
-pytest.importorskip("pandas", minversion="1.1.3")
 pytest.importorskip("laspy", minversion="2")
 
 
@@ -51,7 +50,7 @@ def dataset(
         transforms = nn.Identity()
         return IDTReeS(root, split, task, transforms, download=True, checksum=True)
 
-    @pytest.fixture(params=["pandas", "laspy", "pyvista"])
+    @pytest.fixture(params=["laspy", "pyvista"])
     def mock_missing_module(self, monkeypatch: MonkeyPatch, request: SubRequest) -> str:
         import_orig = builtins.__import__
         package = str(request.param)
@@ -110,13 +109,13 @@ def test_mock_missing_module(
     ) -> None:
         package = mock_missing_module
 
-        if package in ["pandas", "laspy"]:
+        if package == "laspy":
             with pytest.raises(
                 ImportError,
                 match=f"{package} is not installed and is required to use this dataset",
             ):
                 IDTReeS(dataset.root, dataset.split, dataset.task)
-        elif package in ["pyvista"]:
+        elif package == "pyvista":
             with pytest.raises(
                 ImportError,
                 match=f"{package} is not installed and is required to plot point cloud",

tests/datasets/test_inaturalist.py

@@ -1,13 +1,10 @@
 # Copyright (c) Microsoft Corporation. All rights reserved.
 # Licensed under the MIT License.
 
-import builtins
 import os
 from pathlib import Path
-from typing import Any
 
 import pytest
-from pytest import MonkeyPatch
 
 from torchgeo.datasets import (
     BoundingBox,
@@ -16,8 +13,6 @@
     UnionDataset,
 )
 
-pytest.importorskip("pandas", minversion="1.1.3")
-
 
 class TestINaturalist:
     @pytest.fixture(scope="class")
@@ -44,26 +39,6 @@ def test_no_data(self, tmp_path: Path) -> None:
         with pytest.raises(FileNotFoundError, match="Dataset not found"):
             INaturalist(str(tmp_path))
 
-    @pytest.fixture
-    def mock_missing_module(self, monkeypatch: MonkeyPatch) -> None:
-        import_orig = builtins.__import__
-
-        def mocked_import(name: str, *args: Any, **kwargs: Any) -> Any:
-            if name == "pandas":
-                raise ImportError()
-            return import_orig(name, *args, **kwargs)
-
-        monkeypatch.setattr(builtins, "__import__", mocked_import)
-
-    def test_mock_missing_module(
-        self, dataset: INaturalist, mock_missing_module: None
-    ) -> None:
-        with pytest.raises(
-            ImportError,
-            match="pandas is not installed and is required to use this dataset",
-        ):
-            INaturalist(dataset.root)
-
     def test_invalid_query(self, dataset: INaturalist) -> None:
         query = BoundingBox(0, 0, 0, 0, 0, 0)
         with pytest.raises(

tests/datasets/test_openbuildings.py

@@ -1,18 +1,16 @@
 # Copyright (c) Microsoft Corporation. All rights reserved.
 # Licensed under the MIT License.
 
-import builtins
 import json
 import os
 import shutil
 from pathlib import Path
-from typing import Any
 
 import matplotlib.pyplot as plt
+import pandas as pd
 import pytest
 import torch
 import torch.nn as nn
-from _pytest.fixtures import SubRequest
 from pytest import MonkeyPatch
 from rasterio.crs import CRS
 
@@ -23,8 +21,6 @@
     UnionDataset,
 )
 
-pd = pytest.importorskip("pandas", minversion="1.1.3")
-
 
 class TestOpenBuildings:
     @pytest.fixture
@@ -43,30 +39,6 @@ def dataset(self, monkeypatch: MonkeyPatch, tmp_path: Path) -> OpenBuildings:
         transforms = nn.Identity()
         return OpenBuildings(root=root, transforms=transforms)
 
-    @pytest.fixture(params=["pandas"])
-    def mock_missing_module(self, monkeypatch: MonkeyPatch, request: SubRequest) -> str:
-        import_orig = builtins.__import__
-        package = str(request.param)
-
-        def mocked_import(name: str, *args: Any, **kwargs: Any) -> Any:
-            if name == package:
-                raise ImportError()
-            return import_orig(name, *args, **kwargs)
-
-        monkeypatch.setattr(builtins, "__import__", mocked_import)
-        return package
-
-    def test_mock_missing_module(
-        self, dataset: OpenBuildings, mock_missing_module: str
-    ) -> None:
-        package = mock_missing_module
-
-        with pytest.raises(
-            ImportError,
-            match=f"{package} is not installed and is required to use this dataset",
-        ):
-            OpenBuildings(root=dataset.root)
-
     def test_no_shapes_to_rasterize(
         self, dataset: OpenBuildings, tmp_path: Path
     ) -> None:

tests/datasets/test_reforestree.py

@@ -1,11 +1,9 @@
 # Copyright (c) Microsoft Corporation. All rights reserved.
 # Licensed under the MIT License.
 
-import builtins
 import os
 import shutil
 from pathlib import Path
-from typing import Any
 
 import matplotlib.pyplot as plt
 import pytest
@@ -24,7 +22,6 @@ def download_url(url: str, root: str, *args: str) -> None:
 class TestReforesTree:
     @pytest.fixture
     def dataset(self, monkeypatch: MonkeyPatch, tmp_path: Path) -> ReforesTree:
-        pytest.importorskip("pandas", minversion="1.1.3")
         monkeypatch.setattr(torchgeo.datasets.utils, "download_url", download_url)
         data_dir = os.path.join("tests", "data", "reforestree")
 
@@ -54,32 +51,10 @@ def test_getitem(self, dataset: ReforesTree) -> None:
         assert x["image"].ndim == 3
         assert len(x["boxes"]) == 2
 
-    @pytest.fixture
-    def mock_missing_module(self, monkeypatch: MonkeyPatch) -> None:
-        import_orig = builtins.__import__
-        package = "pandas"
-
-        def mocked_import(name: str, *args: Any, **kwargs: Any) -> Any:
-            if name == package:
-                raise ImportError()
-            return import_orig(name, *args, **kwargs)
-
-        monkeypatch.setattr(builtins, "__import__", mocked_import)
-
-    def test_mock_missing_module(
-        self, dataset: ReforesTree, mock_missing_module: None
-    ) -> None:
-        with pytest.raises(
-            ImportError,
-            match="pandas is not installed and is required to use this dataset",
-        ):
-            ReforesTree(root=dataset.root)
-
     def test_len(self, dataset: ReforesTree) -> None:
         assert len(dataset) == 2
 
     def test_not_extracted(self, tmp_path: Path) -> None:
-        pytest.importorskip("pandas", minversion="1.1.3")
         url = os.path.join("tests", "data", "reforestree", "reforesTree.zip")
         shutil.copy(url, tmp_path)
         ReforesTree(root=str(tmp_path))

tests/datasets/test_seasonet.py

@@ -1,12 +1,10 @@
 # Copyright (c) Microsoft Corporation. All rights reserved.
 # Licensed under the MIT License.
 
-import builtins
 import glob
 import os
 import shutil
 from pathlib import Path
-from typing import Any
 
 import matplotlib.pyplot as plt
 import pytest
@@ -19,8 +17,6 @@
 import torchgeo.datasets.utils
 from torchgeo.datasets import SeasoNet
 
-pytest.importorskip("pandas", minversion="1.1.3")
-
 
 def download_url(url: str, root: str, md5: str, *args: str, **kwargs: str) -> None:
     shutil.copy(url, root)
@@ -154,25 +150,6 @@ def test_not_downloaded(self, tmp_path: Path) -> None:
         with pytest.raises(RuntimeError, match="not found in"):
             SeasoNet(str(tmp_path), download=False)
 
-    @pytest.fixture
-    def mock_missing_module(self, monkeypatch: MonkeyPatch) -> None:
-        import_orig = builtins.__import__
-
-        def mocked_import(name: str, *args: Any, **kwargs: Any) -> Any:
-            if name == "pandas":
-                raise ImportError()
-            return import_orig(name, *args, **kwargs)
-
-        monkeypatch.setattr(builtins, "__import__", mocked_import)
-
-    def test_mock_missing_module(
-        self, dataset: SeasoNet, mock_missing_module: None
-    ) -> None:
-        with pytest.raises(
-            ImportError, match="pandas is not installed and is required"
-        ):
-            SeasoNet(dataset.root)
-
     def test_out_of_bounds(self, dataset: SeasoNet) -> None:
         with pytest.raises(IndexError):
             dataset[5]