-
Notifications
You must be signed in to change notification settings - Fork 2.1k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Speed issue with binding from Numpy array to std::vector<Eigen::Vector3d>. #1481
Comments
This issue is highly critical for us at the moment. |
Hi @syncle, could you bisect this to a specific commit? This would make our investigation much easier. Thanks, Wenzel |
Sure. This is a summary:
This is a full log:
|
Hello. Could somebody follow-up on this? I would be appreciated. :) |
We got a great feedback from @wjakob, and we were able to improve performance a lot. Here is a detailed description written by @yxlao in our Open3D repository. The key idea was
For example we defined new functions like below.
Thanks for the feedback and help! |
Thanks for the great library!
I have been using pybind for our open source project. One of the useful feature we are using is binding Numpy array to
std::vector<Eigen::Vector3d>
. For example,in
test_stl_eigen()
function,y
getsstd::vector<Eigen::Vector3d>
.The issue I would like to mention is related to the time. We have been using pybind since 2.0_dev version. With pybind 2.0 dev,
test_stl_eigen
took about 0.3s. With stable release pybind 2.2 or higher,test_stl_eigen
took about 2.3s. Could I ask some help from pybind community?For your reference, here is the cpp code we wrote for binding.
The text was updated successfully, but these errors were encountered: