[REVIEW] Proposal for "proper" C/C++ API (issue #92)#247

Merged

dantegd merged 27 commits intorapidsai:branch-0.6from jirikraus:fea-ext-cuML-cpp-iface

Mar 12, 2019

+1,838-6

Commits on Mar 8, 2019

Proposal for "proper" C/C++ API (issue rapidsai#92 )
jirikraus
committed
Addressed teju85's comments on cuML/DEVELOPER_GUIDE.md
jirikraus
committed
Added cusparseHandle_t to cumlHandle_impl need by UMAP
jirikraus
committed
Fixed comments and used nullptr instead of 0.
jirikraus
committed
Added stream to host_buffer and device_buffer to avoid the default stream and enable stream ordered RAII.
jirikraus
committed
Fixed include of std::vector.
jirikraus
committed
Added Multi GPU section to developer guide proposal.
jirikraus
committed
Switched from typedef to using.
jirikraus
committed
Moved allocator interface to ml-prims
jirikraus
committed
Added thrust execution policy convience function (idea from RMM)
jirikraus
committed
Added internal streams to cumlHandle
jirikraus
committed
Streamlined implementation of resize in device_buffer and host_buffer
jirikraus
committed
Change namespace of deviceAllocator and hostAllocator from ML to MLCommon
jirikraus
committed
Switched to using const ML::cumlHandle_impl& instead of ML::cumlHandle*
jirikraus
committed
Added some content related to thread safety.
jirikraus
committed
Moved device_buffer and host_buffer to ml-prims
jirikraus
committed
Fixed typo in DEVELOPER_GUIDE.md and added cross reference to CONTRIBUTING.md
jirikraus
committed
Fixed DEVELOPER_GUIDE.md formatting and added C APIs section.
jirikraus
committed
Extended C API proposal.
jirikraus
committed
Added buffer_base as base class for device_buffer and host_buffer to avoid code dublication.
jirikraus
committed
Dropped multi GPU part of cumlHandle_impl
jirikraus
committed
Added allocator adapter for STL containers.
jirikraus
committed
Update cuML/DEVELOPER_GUIDE.md

harrism
and
jirikraus
committed
ML::exec_policy -> ML::thrust_exec_policy
jirikraus
committed
Updated cuML/DEVELOPER_GUIDE.md based on feedback from Mark Harris.
jirikraus
committed
Added first version of doxygen documentation to the public part of the interface.
jirikraus
committed

Commits on Mar 12, 2019

Merge branch 'branch-0.6' into fea-ext-cuML-cpp-iface
dantegd
authored

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

[REVIEW] Proposal for "proper" C/C++ API (issue #92)#247

Commits on Mar 8, 2019

Proposal for "proper" C/C++ API (issue rapidsai#92 )

Addressed teju85's comments on cuML/DEVELOPER_GUIDE.md

Added cusparseHandle_t to cumlHandle_impl need by UMAP

Fixed comments and used nullptr instead of 0.

Added stream to host_buffer and device_buffer to avoid the default stream and enable stream ordered RAII.

Fixed include of std::vector.

Added Multi GPU section to developer guide proposal.

Switched from typedef to using.

Moved allocator interface to ml-prims

Added thrust execution policy convience function (idea from RMM)

Added internal streams to cumlHandle

Streamlined implementation of resize in device_buffer and host_buffer

Change namespace of deviceAllocator and hostAllocator from ML to MLCommon

Switched to using const ML::cumlHandle_impl& instead of ML::cumlHandle*

Added some content related to thread safety.

Moved device_buffer and host_buffer to ml-prims

Fixed typo in DEVELOPER_GUIDE.md and added cross reference to CONTRIBUTING.md

Fixed DEVELOPER_GUIDE.md formatting and added C APIs section.

Extended C API proposal.

Added buffer_base as base class for device_buffer and host_buffer to avoid code dublication.

Dropped multi GPU part of cumlHandle_impl

Added allocator adapter for STL containers.

Update cuML/DEVELOPER_GUIDE.md

ML::exec_policy -> ML::thrust_exec_policy

Updated cuML/DEVELOPER_GUIDE.md based on feedback from Mark Harris.

Added first version of doxygen documentation to the public part of the interface.

Commits on Mar 12, 2019

Merge branch 'branch-0.6' into fea-ext-cuML-cpp-iface