numpyeigen

numpyeigen copied to clipboard

Add python scripts to call codegen stuff

2

Let's have scripts that do what cmake does. This makes it easy to integrate numpyeigen into other build systems.

fwilliams

error: ‘class pybind11::dtype’ has no member named ‘type’; did you mean ‘dtype’?

3

My ENV: - OS: Ubuntu 20.04 - Python: 3.8.10 ```console [ 97%] Linking CXX shared library numpyeigen_helpers.cpython-38-x86_64-linux-gnu.so cd ....../numpyeigen/build/tests && /usr/bin/cmake -E cmake_link_script CMakeFiles/numpyeigen_helpers.dir/link.txt --verbose=1 /usr/local/bin/c++ -fPIC -O3 -DNDEBUG -shared...

jiapei100

My ENV: - OS: Ubuntu 20.04.2 - Python: 3.8.10 - pip: 21.2.3 ```console ....../numpyeigen/cmake/../src/npe_sparse_array.h:169:13: error: lvalue required as left operand of assignment 169 | data.flags() &= ~detail::npy_api::NPY_ARRAY_OWNDATA_; | ~~~~~~~~~~^~ ....../numpyeigen/cmake/../src/npe_sparse_array.h:172:15:...

jiapei100

Fix "permission denied" due to parallel builds in Visual Studio (Window 10). I suggest the cross-platform solution from https://stackoverflow.com/a/58955530.

nodtem66

I prefer using manually installed **Eigen3** and **pybind11**, instead of let **numpyeigen** download both packages to folder **exterenal**. Any suggestions please? Cheers

jiapei100

Input arguments are getting reordered by the compiler. Last stable commit seems to be `05a22ad4f06a431bffc3860b30c00cf89e81373b`

fwilliams

missing npe_dense_like

1

https://github.com/geometryprocessing/libigl-python-bindings/issues/28

teseoch

teseoch

Let's add support for ``` npe_class(PythonClassName, cpp::MyClass) npe_init() npe_arg(a, dense_f32, dense_f64) npe_arg(b, npe_matches(a)) npe_begin_code() return cpp::MyClass(a, b); npe_end_code() npe_method(foo) npe_arg(a, dense_f32, dense_f64) npe_begin_code() return npe_self.foo(a); npe_end_code() npe_end_class() ``` This is...

fwilliams

Contribute zero-copy sparse matrices to pybind11?

1

Hi @fwilliams ! This project looks really cool, and it looks like it has (or will have?) a feature that I've been looking for: zero-copy binding of scipy sparse matrices...

matthewwardrop