1

I would like to create a numpy.darray exactly at the location where a ctypes.c_void_p pointer is pointing to.

buffer_ = ctypes.c_void_p()
# magically let the API point it to a special memory location
numpy_buffer_ = numpy.ndarray(shape, dtype=dtype, buffer=buffer_)

However, I am getting the error:

TypeError: buffer is too small for requested array

I got the general idea to solve this problem like this from another thread, where it is used slightly differently:

numpy.ndarray(shape, dtype=dtype, buffer=c_void_p.from_address(address))

However, as I already have a c_void_p pointing to the correct address, I hoped I could use it.

Any ideas, why my buffer appears too small? (It also does not work, if I put a tiny shape to my ndarray)

Best,

gbrown

Improve this question
9
  • 1
    Is your buffer initialized with a size different from 0? I don't know if you really have the choice but this is something that could cause your problem Commented Sep 25, 2018 at 9:10
  • My buffer is a ctypes.c_void_p(), so I do not think it has a size. However, in the example from the other thread, he is using a c_void_p as well, so I would expect that it must work somehow. Commented Sep 25, 2018 at 9:11
  • > it must work somehow. Yeah sure, but why don't you try to follow a working example first before trying to do anything else? I understand that you try by your side but maybe that the difference of your c_void_p initialization makes the difference with the one from the example. Have you read the documentation? Are you sure those two ways produce the exact same result? I don't want to stop you in your learning initiative but developpers write documentation and exemples for a reason.. Commented Sep 25, 2018 at 9:24
  • 2
    numpy.ctypeslib.as_array(obj, shape=None) stackoverflow.com/questions/23930671/… Commented Sep 25, 2018 at 9:31
  • 1
    @Joe: That seems to solve the problem. Sorry, I must have missed this during my search! Thanks to the both of you! Commented Sep 25, 2018 at 9:42
Related questions
19
How to use NumPy array with ctypes?
44
Getting data from ctypes array into numpy
7
ctypes pointer into the middle of a numpy array
Load 7 more related questions Show fewer related questions

0

Reset to default

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Start asking to get answers

Find the answer to your question by asking.

Ask question

Explore related questions

See similar questions with these tags.