Numpy does not have float128 on all platforms

Author: oleksandr-pavlyk

mkl_fft/_float_utils.py

@@ -24,8 +24,7 @@
 # OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
 # OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
 
-from numpy import (half, float32, asarray, ndarray,
-                   longdouble, float64, longcomplex, complex_, float128, complex256)
+import numpy as np
 
 __all__ = ['__upcast_float16_array', '__downcast_float128_array', '__supported_array_or_not_implemented']
 
@@ -35,18 +34,18 @@ def __upcast_float16_array(x):
     instead of float64, as mkl_fft would do"""
     if hasattr(x, "dtype"):
         xdt = x.dtype
-        if xdt == half:
+        if xdt == np.half:
             # no half-precision routines, so convert to single precision
-            return asarray(x, dtype=float32)
-        if xdt == longdouble and not xdt == float64:
+            return np.asarray(x, dtype=np.float32)
+        if xdt == np.longdouble and not xdt == np.float64:
             raise ValueError("type %s is not supported" % xdt)
-    if not isinstance(x, ndarray):
-        __x = asarray(x)
+    if not isinstance(x, np.ndarray):
+        __x = np.asarray(x)
         xdt = __x.dtype
-        if xdt == half:
+        if xdt == np.half:
             # no half-precision routines, so convert to single precision
-            return asarray(__x, dtype=float32)
-        if xdt == longdouble and not xdt == float64:
+            return np.asarray(__x, dtype=np.float32)
+        if xdt == np.longdouble and not xdt == np.float64:
             raise ValueError("type %s is not supported" % xdt)
         return __x
     return x
@@ -58,17 +57,17 @@ def __downcast_float128_array(x):
     complex128, instead of raising an error"""
     if hasattr(x, "dtype"):
         xdt = x.dtype
-        if xdt == longdouble and not xdt == float64:
-            return asarray(x, dtype=float64)
-        elif xdt == longcomplex and not xdt == complex_:
-            return asarray(x, dtype=complex_)
-    if not isinstance(x, ndarray):
-        __x = asarray(x)
+        if xdt == np.longdouble and not xdt == np.float64:
+            return np.asarray(x, dtype=np.float64)
+        elif xdt == np.longcomplex and not xdt == np.complex_:
+            return np.asarray(x, dtype=np.complex_)
+    if not isinstance(x, np.ndarray):
+        __x = np.asarray(x)
         xdt = __x.dtype
-        if xdt == longdouble and not xdt == float64:
-            return asarray(x, dtype=float64)
-        elif xdt == longcomplex and not xdt == complex_:
-            return asarray(x, dtype=complex_)
+        if xdt == np.longdouble and not xdt == np.float64:
+            return np.asarray(x, dtype=np.float64)
+        elif xdt == np.longcomplex and not xdt == np.complex_:
+            return np.asarray(x, dtype=np.complex_)
         return __x
     return x
 
@@ -78,7 +77,12 @@ def __supported_array_or_not_implemented(x):
     Used in _scipy_fft_backend to convert array to float32,
     float64, complex64, or complex128 type or return NotImplemented
     """
-    __x = asarray(x)
-    if __x.dtype in [half, float128, complex256]:
+    __x = np.asarray(x)
+    black_list = [np.half]
+    if hasattr(np, 'float128'):
+        black_list.append(np.float128)
+    if hasattr(np, 'complex256'):
+        black_list.append(np.complex256)
+    if __x.dtype in black_list:
         return NotImplemented
     return __x