Update tests

test/__init__.py

@@ -0,0 +1,3 @@
+# Copyright (C) 2024 Roberto Rossini <[email protected]>
+#
+# SPDX-License-Identifier: MIT

test/helpers.py

@@ -0,0 +1,39 @@
+# Copyright (C) 2024 Roberto Rossini <[email protected]>
+#
+# SPDX-License-Identifier: MIT
+
+
+def numpy_avail() -> bool:
+    try:
+        import numpy
+    except ModuleNotFoundError:
+        return False
+
+    return True
+
+
+def pandas_avail() -> bool:
+    try:
+        import pandas
+    except ModuleNotFoundError:
+        return False
+
+    return True
+
+
+def pyarrow_avail() -> bool:
+    try:
+        import pyarrow
+    except ModuleNotFoundError:
+        return False
+
+    return True
+
+
+def scipy_avail() -> bool:
+    try:
+        import scipy
+    except ModuleNotFoundError:
+        return False
+
+    return True

test/test_fetch_dense.py

@@ -4,12 +4,12 @@
 
 import pathlib
 
-import numpy as np
-import numpy.typing as npt
 import pytest
 
 import hictkpy
 
+from .helpers import numpy_avail
+
 testdir = pathlib.Path(__file__).resolve().parent
 
 pytestmark = pytest.mark.parametrize(
@@ -21,7 +21,9 @@
 )
 
 
-def issymmetric(m: npt.NDArray) -> bool:
+def issymmetric(m) -> bool:
+    import numpy as np
+
     assert m.ndim == 2
     if m.size == 0:
         return True
@@ -31,6 +33,7 @@ def issymmetric(m: npt.NDArray) -> bool:
     return np.allclose(m, m.T, atol=0.0, rtol=0.0)
 
 
+@pytest.mark.skipif(not numpy_avail(), reason="numpy is not available")
 class TestClass:
     def test_genome_wide(self, file, resolution):
         f = hictkpy.File(file, resolution)
@@ -40,6 +43,8 @@ def test_genome_wide(self, file, resolution):
         assert issymmetric(m)
 
     def test_cis(self, file, resolution):
+        import numpy as np
+
         f = hictkpy.File(file, resolution)
         m = f.fetch("chr2R:10,000,000-15,000,000").to_numpy()
         assert m.shape == (50, 50)
@@ -68,6 +73,8 @@ def test_cis(self, file, resolution):
         assert m.sum() == 12_607_205
 
     def test_trans(self, file, resolution):
+        import numpy as np
+
         f = hictkpy.File(file, resolution)
         m = f.fetch("chr2R:10,000,000-15,000,000", "chrX:0-10,000,000").to_numpy()
         assert m.shape == (50, 100)

test/test_fetch_df.py

@@ -2,13 +2,15 @@
 #
 # SPDX-License-Identifier: MIT
 
+import math
 import pathlib
 
-import numpy as np
 import pytest
 
 import hictkpy
 
+from .helpers import numpy_avail, pandas_avail, pyarrow_avail
+
 testdir = pathlib.Path(__file__).resolve().parent
 
 pytestmark = pytest.mark.parametrize(
@@ -20,16 +22,7 @@
 )
 
 
-def pandas_avail() -> bool:
-    try:
-        import pandas
-    except ModuleNotFoundError:
-        return False
-
-    return True
-
-
-@pytest.mark.skipif(not pandas_avail(), reason="pandas is not available")
+@pytest.mark.skipif(not pandas_avail() or not pyarrow_avail(), reason="either pandas or pyarrow are not available")
 class TestClass:
     def test_genome_wide(self, file, resolution):
         f = hictkpy.File(file, resolution)
@@ -54,11 +47,14 @@ def test_cis(self, file, resolution):
         assert df["chrom1"].dtype.name == "category"
         assert df["chrom2"].dtype.name == "category"
 
-        df = f.fetch("chr2R:10,000,000-15,000,000", count_type="int").to_df()
-        assert df["count"].dtype == np.int32
+        if numpy_avail():
+            import numpy as np
+
+            df = f.fetch("chr2R:10,000,000-15,000,000", count_type="int").to_df()
+            assert df["count"].dtype == np.int32
 
-        df = f.fetch("chr2R:10,000,000-15,000,000", count_type="float").to_df()
-        assert df["count"].dtype == np.float64
+            df = f.fetch("chr2R:10,000,000-15,000,000", count_type="float").to_df()
+            assert df["count"].dtype == np.float64
 
         df = f.fetch("chr2R\t10000000\t15000000", query_type="BED").to_df()
         assert len(df) == 1275
@@ -87,11 +83,14 @@ def test_trans(self, file, resolution):
         assert df["chrom1"].dtype.name == "category"
         assert df["chrom2"].dtype.name == "category"
 
-        df = f.fetch("chr2R:10,000,000-15,000,000", "chrX:0-10,000,000", count_type="int").to_df()
-        assert df["count"].dtype == np.int32
+        if numpy_avail():
+            import numpy as np
+
+            df = f.fetch("chr2R:10,000,000-15,000,000", "chrX:0-10,000,000", count_type="int").to_df()
+            assert df["count"].dtype == np.int32
 
-        df = f.fetch("chr2R:10,000,000-15,000,000", "chrX:0-10,000,000", count_type="float").to_df()
-        assert df["count"].dtype == np.float64
+            df = f.fetch("chr2R:10,000,000-15,000,000", "chrX:0-10,000,000", count_type="float").to_df()
+            assert df["count"].dtype == np.float64
 
         df = f.fetch("chr2R\t10000000\t15000000", "chrX\t0\t10000000", query_type="BED").to_df()
         assert len(df) == 4995
@@ -112,4 +111,4 @@ def test_balanced(self, file, resolution):
         else:
             df = f.fetch("chr2R:10,000,000-15,000,000", normalization="ICE").to_df()
 
-        assert np.isclose(59.349524704033215, df["count"].sum())
+        assert math.isclose(59.349524704033215, df["count"].sum(), rel_tol=1.0e-5, abs_tol=1.0e-8)

test/test_fetch_iters.py

@@ -2,9 +2,9 @@
 #
 # SPDX-License-Identifier: MIT
 
+import math
 import pathlib
 
-import numpy as np
 import pytest
 
 import hictkpy
@@ -87,4 +87,4 @@ def test_balanced(self, file, resolution):
         else:
             sel = f.fetch("chr2R:10,000,000-15,000,000", normalization="ICE")
 
-        assert np.isclose(59.349524704033215, compute_sum(sel))
+        assert math.isclose(59.349524704033215, compute_sum(sel), rel_tol=1.0e-5, abs_tol=1.0e-8)

test/test_fetch_sparse.py

@@ -2,13 +2,15 @@
 #
 # SPDX-License-Identifier: MIT
 
+import math
 import pathlib
 
-import numpy as np
 import pytest
 
 import hictkpy
 
+from .helpers import numpy_avail, scipy_avail
+
 testdir = pathlib.Path(__file__).resolve().parent
 
 pytestmark = pytest.mark.parametrize(
@@ -20,15 +22,6 @@
 )
 
 
-def scipy_avail() -> bool:
-    try:
-        import scipy
-    except ModuleNotFoundError:
-        return False
-
-    return True
-
-
 @pytest.mark.skipif(not scipy_avail(), reason="scipy is not available")
 class TestClass:
     def test_genome_wide(self, file, resolution):
@@ -45,11 +38,14 @@ def test_cis(self, file, resolution):
         assert m.shape == (50, 50)
         assert m.sum() == 4_519_080
 
-        m = f.fetch("chr2R:10,000,000-15,000,000", count_type="int").to_coo()
-        assert m.dtype == np.int32
+        if numpy_avail():
+            import numpy as np
+
+            m = f.fetch("chr2R:10,000,000-15,000,000", count_type="int").to_coo()
+            assert m.dtype == np.int32
 
-        m = f.fetch("chr2R:10,000,000-15,000,000", count_type="float").to_coo()
-        assert m.dtype == np.float64
+            m = f.fetch("chr2R:10,000,000-15,000,000", count_type="float").to_coo()
+            assert m.dtype == np.float64
 
         m = f.fetch("chr2R\t10000000\t15000000", query_type="BED").to_coo()
         assert m.shape == (50, 50)
@@ -69,11 +65,14 @@ def test_trans(self, file, resolution):
         assert m.shape == (50, 100)
         assert m.sum() == 83_604
 
-        m = f.fetch("chr2R:10,000,000-15,000,000", "chrX:0-10,000,000", count_type="int").to_coo()
-        assert m.dtype == np.int32
+        if numpy_avail():
+            import numpy as np
+
+            m = f.fetch("chr2R:10,000,000-15,000,000", "chrX:0-10,000,000", count_type="int").to_coo()
+            assert m.dtype == np.int32
 
-        m = f.fetch("chr2R:10,000,000-15,000,000", "chrX:0-10,000,000", count_type="float").to_coo()
-        assert m.dtype == np.float64
+            m = f.fetch("chr2R:10,000,000-15,000,000", "chrX:0-10,000,000", count_type="float").to_coo()
+            assert m.dtype == np.float64
 
         m = f.fetch("chr2R\t10000000\t15000000", "chrX\t0\t10000000", query_type="BED").to_coo()
         assert m.shape == (50, 100)
@@ -86,4 +85,4 @@ def test_balanced(self, file, resolution):
         else:
             m = f.fetch("chr2R:10,000,000-15,000,000", normalization="ICE").to_coo()
 
-        assert np.isclose(59.349524704033215, m.sum())
+        assert math.isclose(59.349524704033215, m.sum(), rel_tol=1.0e-5, abs_tol=1.0e-8)

test/test_file_creation_cool.py

@@ -30,7 +30,16 @@ def pandas_avail() -> bool:
     return True
 
 
-@pytest.mark.skipif(not pandas_avail(), reason="pandas is not available")
+def pyarrow_avail() -> bool:
+    try:
+        import pyarrow
+    except ModuleNotFoundError:
+        return False
+
+    return True
+
+
+@pytest.mark.skipif(not pandas_avail() or not pyarrow_avail(), reason="either pandas or pyarrow are not available")
 class TestClass:
     @staticmethod
     def setup_method():

test/test_file_creation_hic.py

@@ -30,7 +30,16 @@ def pandas_avail() -> bool:
     return True
 
 
-@pytest.mark.skipif(not pandas_avail(), reason="pandas is not available")
+def pyarrow_avail() -> bool:
+    try:
+        import pyarrow
+    except ModuleNotFoundError:
+        return False
+
+    return True
+
+
+@pytest.mark.skipif(not pandas_avail() or not pyarrow_avail(), reason="either pandas or pyarrow are not available")
 class TestClass:
     @staticmethod
     def setup_method():