All Questions
Tagged with cython memoryview
83 questions
1
vote
1
answer
81
views
Cython fused dtype memoryview won't complie
I have some code where I (am trying to) do roughly the following:
ctypedef fused Raster_t:
numpy.uint8_t
numpy.uint16_t
cdef class MyClass:
# shape is (rows, cols, channels)
cdef ...
0
votes
1
answer
39
views
Cython: How to pass a numpy memoryview to a nogil cdef function
I would like to send a numpy array (memoryview) to another Cython function.
cdef int function(int[:,:] ref) nogil:
pass
cpdef test():
cdef np.ndarray[int, ndim=2] ref = np.full((10,10),0)
...
0
votes
1
answer
70
views
Creating a cython memoryview manually
I have a contiguous 1-d numpy array that holds data. However, this array is really a buffer, and I have several numpy views to access that data. These views have different shapes, ndims, offsets etc .....
- 461
1
vote
0
answers
109
views
Cython: Memoryviewslice vs C array
I define a C function with memoryview input to work with a NumPy array, but a pure C-defined temporary float array can not work with 'base_func'. Error:
Operation not allowed without gil
How can I ...
- 45
2
votes
1
answer
206
views
Cython returned memoryview is always considered uninitialized
Similar to Cython Memoryview as return value, but I didn't see a solution other than hacking the generated C code.
I'm using Cython 3.0, but it looks like the result is the same with <3.
Here's an ...
- 3,647
0
votes
0
answers
78
views
How to transpose a Cython memoryview slice?
I am trying to calculate the transpose of a Cython typed memoryview slice (eg. a 2D part of a 3D array). I am encountering a transpose error specifically when I try to transpose a 2D slice of a 3D ...
- 35
1
vote
2
answers
120
views
Why does the memoryview assignment from array have python interactions?
When compiled with cython -a my-file.pyx, this simple line of cdef is annotated as yellow in the html file.
# my-file.pyx
from cpython.array cimport array
def f(double[:] xyz):
cdef double[:] ...
- 20.8k
1
vote
1
answer
62
views
Lackluster performance using inlined function in cython memoryview
I wrote the following Cython code that calculates pairwise distances à la scipy.spatial.distance.cdist.
# cython: infer_types=True
# cython: boundscheck=False
# cython: wraparound=False
cimport numpy ...
- 717
2
votes
1
answer
417
views
Maximizing the speed of Cython array operations when using typed memory views and BLAS
I'm trying to maximize the speed of my Cython 3.0 code that involves updating an array with a loop of several array operations (including matrix-vector multiplication, vector-vector addition, and ...
- 35
0
votes
1
answer
90
views
Cython: How to convert numpy 2D array of type "object" to memoryview?
I am wondering how one can convert a 2D numpy array input of type "object" into a Cython memoryview?
For example, if I have the following Cython extension type:
cdef class A:
cdef:
...
- 1,252
1
vote
1
answer
96
views
Cython: safe way to check whether element in malloc memoryview is assigned
I want to create various small memoryviews in function temporarily and release them when the function is finished. Seems initialize the memoryview by malloc is better, based on the comparisons ...
- 606
2
votes
0
answers
1k
views
Cython vs. Pure Python with Memoryviews
I'm just starting to learn Cython and I'm confused on an issue. If you look at the main tutorial they have examples that show a "Pure Python" mode vs. "Cython" mode. The pure ...
- 376
2
votes
1
answer
284
views
Cython: Casting vector of C++ `std::complex<double>` to C `double complex`
In Cython, starting from a C++ vector of std::complex<double> like this...
# Imports
from libcpp.vector cimport vector
from libcpp.complex cimport complex as cpp_complex
ctypedef my_complex = ...
- 7,491
1
vote
1
answer
216
views
Cython memoryview shape incorrect?
Consider the following to create a linear array of size 4:
import numpy as np
cimport numpy as np
cdef np.float64_t [:] a = np.zeros(shape=(4),dtype=np.float64)
a.shape should be (4,). However:
print(...
- 661
0
votes
1
answer
91
views
Cython pyx modification for multidimensional numpy input with memory views
My current input of a function wrapped from a c++ class is a one dimensional numpy array, e.g., func(np.array([1,2,3])) for which my cython description in the pyx looks like the following:
def func(...
- 949