Hacktoberfest 2026: the issues maintainers tagged for October, open and beginner-friendly. Browse Hacktoberfest issues

inconsistent numpy dtype in python and C++

Open
#304 0 comments 0 reactions 0 assignees View on GitHub

Nobody has claimed this yet.

Assessment

Difficulty
4/5
Estimated time
3-5 days
Newbie friendliness
35/100
Issue type
Bug
Clarity
Mostly clear
Activity status
Stale
Tech stack
cpp, numpy, python
Domain
api, backend

Research direction

Reproduce the report using the provided dummy.cpp extension and run_dummpy.py script with the stated Python, NumPy, and Boost versions. Inspect how the two ndarray arguments reach the Boost.NumPy binding and verify that each retains its dtype; done means the C++ output matches float32 and float64 consistently.

Written by the indexing model from the issue text.

Description

Hi,
I'm trying to use the boost numpy bindings. If I pass two ndarray with different dtype to C++, they don't preserve the correct dtype. Here is a simple example:
dummy.cpp:

#include <iostream>
#include <boost/python.hpp>
#include <boost/python/numpy.hpp>

namespace py = boost::python;
namespace np = boost::python::numpy;

void dummy(np::ndarray const& a, np::ndarray const& b) {

    char const* at = py::extract<char const* >(py::str(a.get_dtype()));
    char const* bt = py::extract<char const* >(py::str(b.get_dtype()));

    std::cout << at << std::endl;
    std::cout << bt << std::endl;
}


BOOST_PYTHON_MODULE(dummy) {
   Py_Initialize(); // need initialize in BOOST_PYTHON_MODULE, or will get segmentation fault and core dump 
   np::initialize();
   py::def("dummy", dummy);
}

compile

g++ -lboost_python38 -lboost_numpy38 -L$CONDA_PREFIX/lib -I$CONDA_PREFIX/include -I$CONDA_PREFIX/include/python3.8 -fPIC --shared -o dummy.so dummy.cpp

run_dummpy.py

import numpy as np
import dummy

a = np.array([[1, 2, 3], [4, 5, 6]], dtype = np.float32)
print("python", a.dtype)
b = np.array([[1, 2, 3], [4, 5, 6]], dtype = np.double)
print("python", b.dtype)

dummy.dummy(a, b)

The output is inconsistent. The first time I run it by python run_dummy.py it gives

python float32
python float64
__str__
float64

If I re-run it python run_dummy.py , it gives

python float32
python float64
float64
float64

Still not the right output.
I'm using python 3.8.2, numpy 1.18.1, and boost 1.72.0. boost is installed from anaconda conda install -c conda-forge boost
Could you help? Thanks

Dominant language
C++
Stars
537
Forks
223
Avg merge
11h 22m
Merged PRs (30d)
2

Contributor guide

No contributing guide indexed for this repository

First steps

  1. Read the whole issue, then the project's contributing guide.
  2. Comment on the issue to say you are picking it up — it saves two people doing the same work.
  3. Fork the repository and make your change on a branch.
  4. Open a pull request that references the issue number.

More from boostorg/python

All issues in boostorg/python

Similar issues

More C++ issues

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.