cupy.random.bytes not working

• Conditions (you can just paste the output of python -c 'import cupy; cupy.show_config()')

OS                           : Linux-5.9.8-2-MANJARO-x86_64-with-arch-Manjaro-Linux
CuPy Version                 : 8.1.0
NumPy Version                : 1.19.4
SciPy Version                : 1.5.3
CUDA Root                    : /[hiden]/.conda/envs/rapids-0.16
CUDA Build Version           : 10010
CUDA Driver Version          : 11010
CUDA Runtime Version         : 10010
cuBLAS Version               : 10201
cuFFT Version                : 10101
cuRAND Version               : 10101
cuSOLVER Version             : (10, 2, 0)
cuSPARSE Version             : 10300
NVRTC Version                : (10, 1)
Thrust Version               : 100906
CUB Build Version            : 100800
cuDNN Build Version          : None
cuDNN Version                : None
NCCL Build Version           : 2708
NCCL Runtime Version         : 2708
cuTENSOR Version             : None
Device 0 Name                : GeForce GTX 1080 Ti
Device 0 Compute Capability  : 61

• Code to reproduce

import cupy as cp
b = cp.random.bytes(10)

• Error messages, stack traces, or logs

---------------------------------------------------------------------------
AttributeError                            Traceback (most recent call last)
<ipython-input-121-104e7c75af44> in <module>
      1 import cupy as cp
----> 2 b = cp.random.bytes(10)

~/.conda/envs/rapids-0.16/lib/python3.7/site-packages/cupy/random/__init__.py in bytes(length)
      8                  <numpy.random.mtrand.RandomState.bytes>`
      9     """
---> 10     return _numpy.bytes(length)
     11 
     12 

~/.conda/envs/rapids-0.16/lib/python3.7/site-packages/numpy/__init__.py in __getattr__(attr)
    213             else:
    214                 raise AttributeError("module {!r} has no attribute "
--> 215                                      "{!r}".format(__name__, attr))
    216 
    217         def __dir__():

AttributeError: module 'numpy' has no attribute 'bytes'

[lucasrodes]

Thanks for your hints, @kmaehashi, really appreciate it. While trying this I ran into an error. I opened issue #4321

[kmaehashi]

Thanks for sharing the details of your use-case @lucasrodes! In your case, I think it’s better to generate and handle data as a sequence of uint8 instead of bytes.

N = 10
L = 8
data = numpy.random.randint(0,256, size=(N, L), dtype=numpy.uint8)  # or cupy instead of numpy

As for hash functions, you will need to write or find the CUDA implementation like this one, load it using cupy.RawKernel and pass the generated random array to it.