def __mul__(self, other):
if isinstance(other, gpuarray.GPUArray):
result = type(self)(np.broadcast_shapes(self.shape, other.shape),
gpuarray._get_common_dtype(self, other))
return self._elwise_multiply(other, result)
else:
return super().__mul__(other)
def weighted_inner(self, b, w):
if w is None:
return self.inner(b)
a = self.data
dtype_out = _get_common_dtype(a, b)
krnl = get_weighted_inner_kernel(a.dtype, b.dtype, w.dtype, dtype_out)
return krnl(a, b, w).get().max()
def weighted_inner(self, b, w):
if w is None:
return self.inner(other)
a = self.data
dtype_out = _get_common_dtype(a, b)
krnl = get_weighted_inner_kernel(a.dtype, b.dtype, w.dtype, dtype_out)
return krnl(a, b, w).get().max()
def guarded_div(self, a, b):
from pycuda.gpuarray import _get_common_dtype
result = a._new_like_me(_get_common_dtype(a, b))
assert a.shape == b.shape
func = self.guarded_div_kernel(a.dtype, b.dtype, result.dtype)
func.prepared_async_call(a._grid, a._block, None, a.gpudata, b.gpudata,
result.gpudata, a.mem_size)
return result
def guarded_div(self, a, b):
from pycuda.gpuarray import _get_common_dtype
result = a._new_like_me(_get_common_dtype(a, b))
assert a.shape == b.shape
func = self.guarded_div_kernel(a.dtype, b.dtype, result.dtype)
func.prepared_async_call(a._grid, a._block, None,
a.gpudata, b.gpudata,
result.gpudata, a.mem_size)
return result
def __add__(self, other, sub=False, rsub=False):
"""Add an array with an array or an array with a scalar."""
if isinstance(other, gpuarray.GPUArray):
# add another vector
result = type(self)(np.broadcast_shapes(self.shape, other.shape),
gpuarray._get_common_dtype(self, other))
return self._axpbyz(-1 if rsub else 1, other, -1 if sub else 1,
result)
else:
if sub:
return super().__sub__(other)
elif rsub:
return super().__rsub__(other)
else:
return super().__add__(other)
def weighted_inner(a, b, w):
dtype_out = _get_common_dtype(a, b)
krnl = get_weighted_inner_kernel(a.dtype, b.dtype, w.dtype, dtype_out)
return krnl(a, b, w)
def inner(a, b):
dtype_out = _get_common_dtype(a, b)
krnl = get_inner_kernel(a.dtype, b.dtype, dtype_out)
return krnl(a, b)
def __init__(self, x, y, z):
self.x = x.data
self.y = y.data
self.z = z.data
dtype_out = _get_common_dtype(x, y)
self.krnl = get_correlate_kernel(x.dtype, y.dtype, dtype_out)
def correlate(a, b, out, stream=None):
dtype_out = _get_common_dtype(a, b)
krnl = get_correlate_kernel(a.dtype, b.dtype, dtype_out)
krnl(a.data, b.data, out.data)
def inner(self, b):
a = self.data
dtype_out = _get_common_dtype(a, b)
krnl = get_inner_kernel(a.dtype, b.dtype, dtype_out)
return krnl(a, b).get().max()
def inner(self, b):
a = self.data
dtype_out = _get_common_dtype(a, b)
krnl = get_inner_kernel(a.dtype, b.dtype, dtype_out)
return krnl(a, b).get().max()
def __init__(self, x, y, z):
self.x = x.data
self.y = y.data
self.z = z.data
dtype_out = _get_common_dtype(x, y)
self.krnl = get_correlate_kernel(x.dtype, y.dtype, dtype_out)
def correlate(a, b, out, stream=None):
dtype_out = _get_common_dtype(a,b)
krnl = get_correlate_kernel(a.dtype, b.dtype, dtype_out)
krnl(a.data, b.data, out.data)
