Fix svd flip

dask_ml/utils.py

@@ -23,8 +23,18 @@
 logger = logging.getLogger()
 
 
+def _svd_flip_copy(x, y):
+    # If the array is locked, copy the array and transpose it
+    # This happens with a very large array > 1TB
+    # GH: issue 592
+    try:
+        return skm.svd_flip(x, y)
+    except ValueError:
+        return skm.svd_flip(x.copy(), y.copy())
+
+
 def svd_flip(u, v):
-    u2, v2 = delayed(skm.svd_flip, nout=2)(u, v)
+    u2, v2 = delayed(_svd_flip_copy, nout=2)(u, v)
     u = da.from_delayed(u2, shape=u.shape, dtype=u.dtype)
     v = da.from_delayed(v2, shape=v.shape, dtype=v.dtype)
     return u, v

docs/source/changelog.rst

@@ -5,6 +5,7 @@ Version 1.1.1
 ~~~~~~~~~~~~~
 
 - Fixed an issue with the 1.1.0 wheel (:issue:`575`)
+- Make svd_flip work even when arrays are read only (:issue:`592`)
 
 Version 1.1.0
 ~~~~~~~~~~~~~

tests/test_pca.py

@@ -14,7 +14,7 @@
 
 import dask_ml.decomposition as dd
 from dask_ml.decomposition._compat import _assess_dimension_, _infer_dimension_
-from dask_ml.utils import assert_estimator_equal
+from dask_ml.utils import assert_estimator_equal, svd_flip
 
 with warnings.catch_warnings():
     warnings.simplefilter("ignore", FutureWarning)
@@ -782,3 +782,22 @@ def test_pca_sklearn_inputs(input_type, solver):
         a.fit(Y)
     with pytest.raises(TypeError, match="unsupported type"):
         a.fit_transform(Y)
+
+
+def test_svd_flip():
+    rng = np.random.RandomState(0)
+    u = rng.randn(8, 3)
+    v = rng.randn(3, 10)
+    u = da.from_array(u, chunks=(-1, -1))
+    v = da.from_array(v, chunks=(-1, -1))
+    u2, v2 = svd_flip(u, v)
+
+    def set_readonly(x):
+        x.setflags(write=False)
+        return x
+
+    u = u.map_blocks(set_readonly)
+    v = v.map_blocks(set_readonly)
+    u, v = svd_flip(u, v)
+    assert_eq(u, u2)
+    assert_eq(v, v2)