Skip to content

Latest commit

 

History

History
216 lines (189 loc) · 10.1 KB

CHANGELOG.md

File metadata and controls

216 lines (189 loc) · 10.1 KB

cuGraph 0.10.0 (Date TBD)

New Features

Improvements

  • PR #466 Add file splitting test; Update to reduce dask overhead
  • PR #468 Remove unnecessary print statement
  • PR #464 Limit initial RMM pool allocator size to 128mb so pytest can run in parallel

Bug Fixes

  • PR #458 Fix potential race condition in SSSP

cuGraph 0.9.0 (Date TBD)

New Features

  • PR #361 Prototypes for cusort functions
  • PR #357 Pagerank cpp API
  • PR #366 Adds graph.degrees() function returning both in and out degree.
  • PR #380 First implemention of cusort - SNMG key/value sorting
  • PR #416 OpenCypher: Added C++ implementation of db_object class and assorted other classes
  • PR #411 Integrate dask-cugraph in cugraph
  • PR #411 Integrate dask-cugraph in cugraph #411
  • PR #418 Update cusort to handle SNMG key-only sorting
  • PR #423 Add Strongly Connected Components (GEMM); Weakly CC updates;
  • PR #437 Streamline CUDA_REL environment variable
  • PR #449 Fix local build generated file ownerships
  • PR #454 Initial version of updated script to run benchmarks

Improvements

  • PR #353 Change snmg python wrapper in accordance to cpp api
  • PR #362 Restructured python/cython directories and files.
  • PR #365 Updates for setting device and vertex ids for snmg pagerank
  • PR #383 Exposed MG pagerank solver parameters
  • PR #399 Example Prototype of Strongly Connected Components using primitives
  • PR #419 Version test
  • PR #420 drop duplicates, remove print, compute/wait read_csv in pagerank.py
  • PR #439 More efficient computation of number of vertices from edge list
  • PR #445 Update view_edge_list, view_adj_list, and view_transposed_adj_list to return edge weights.
  • PR #450 Add a multi-GPU section in cuGraph documentation.

Bug Fixes

  • PR #368 Bump cudf dependency versions for cugraph conda packages
  • PR #354 Fixed bug in building a debug version
  • PR #360 Fixed bug in snmg coo2csr causing intermittent test failures.
  • PR #364 Fixed bug building or installing cugraph when conda isn't installed
  • PR #375 Added a function to initialize gdf columns in cugraph #375
  • PR #378 cugraph was unable to import device_of_gpu_pointer
  • PR #384 Fixed bug in snmg coo2csr causing error in dask-cugraph tests.
  • PR #382 Disabled vertex id check to allow Azure deployment
  • PR #410 Fixed overflow error in SNMG COO2CSR
  • PR #395 run omp_ge_num_threads in a parallel context
  • PR #412 Fixed formatting issues in cuGraph documentation.
  • PR #413 Updated python build instructions.
  • PR #414 Add weights to wjaccrd.py
  • PR #436 Fix Skip Test Functionality
  • PR #438 Fix versions of packages in build script and conda yml
  • PR #441 Import cudf_cpp.pxd instead of duplicating cudf definitions.
  • PR #441 Removed redundant definitions of python dictionaries and functions.
  • PR #442 Updated versions in conda environments.
  • PR #442 Added except + to cython bindings to C(++) functions.
  • PR #443 Fix accuracy loss issue for snmg pagerank
  • PR #444 Fix warnings in strongly connected components
  • PR #446 Fix permission for source (-x) and script (+x) files.
  • PR #448 Import filter_unreachable
  • PR #453 Re-sync cugraph with cudf (dependencies, type conversion & scatter functions).
  • PR #463 Remove numba dependency and use the one from cudf

cuGraph 0.8.0 (27 June 2019)

New Features

  • PR #287 SNMG power iteration step1
  • PR #297 SNMG degree calculation
  • PR #300 Personalized Page Rank
  • PR #302 SNMG CSR Pagerank (cuda/C++)
  • PR #315 Weakly Connected Components adapted from cuML (cuda/C++)
  • PR #323 Add test skipping function to build.sh
  • PR #308 SNMG python wrapper for pagerank
  • PR #321 Added graph initialization functions for NetworkX compatibility.
  • PR #332 Added C++ support for strings in renumbering function
  • PR #325 Implement SSSP with predecessors (cuda/C++)
  • PR #331 Python bindings and test for Weakly Connected Components.
  • PR #339 SNMG COO2CSR (cuda/C++)
  • PR #341 SSSP with predecessors (python) and function for filtering unreachable nodes in the traversal
  • PR #348 Updated README for release

Improvements

  • PR #291 nvGraph is updated to use RMM instead of directly invoking cnmem functions.
  • PR #286 Reorganized cugraph source directory
  • PR #306 Integrated nvgraph to libcugraph.so (libnvgraph_rapids.so will not be built anymore).
  • PR #306 Updated python test files to run pytest with all four RMM configurations.
  • PR #321 Added check routines for input graph data vertex IDs and offsets (cugraph currently supports only 32-bit integers).
  • PR #333 Various general improvements at the library level

Bug Fixes

  • PR #283 Automerge fix
  • PR #291 Fixed a RMM memory allocation failure due to duplicate copies of cnmem.o
  • PR #291 Fixed a cub CsrMV call error when RMM pool allocator is used.
  • PR #306 Fixed cmake warnings due to library conflicts.
  • PR #311 Fixed bug in SNMG degree causing failure for three gpus
  • PR #309 Update conda build recipes
  • PR #314 Added datasets to gitignore
  • PR #322 Updates to accommodate new cudf include file locations
  • PR #324 Fixed crash in WeakCC for larger graph and added adj matrix symmetry check
  • PR #327 Implemented a temporary fix for the build failure due to gunrock updates.
  • PR #345 Updated CMakeLists.txt to apply RUNPATH to transitive dependencies.
  • PR #350 Configure Sphinx to render params correctly
  • PR #359 Updates to remove libboost_system as a runtime dependency on libcugraph.so

cuGraph 0.7.0 (10 May 2019)

New Features

  • PR #195 Added Graph.get_two_hop_neighbors() method
  • PR #195 Updated Jaccard and Weighted Jaccard to accept lists of vertex pairs to compute for
  • PR #202 Added methods to compute the overlap coefficient and weighted overlap coefficient
  • PR #230 SNMG SPMV and helpers functions
  • PR #210 Expose degree calculation kernel via python API
  • PR #220 Added bindings for Nvgraph triangle counting
  • PR #234 Added bindings for renumbering, modify renumbering to use RMM
  • PR #246 Added bindings for subgraph extraction
  • PR #250 Add local build script to mimic gpuCI
  • PR #261 Add docs build script to cuGraph
  • PR #301 Added build.sh script, updated CI scripts and documentation

Improvements

  • PR #157 Removed cudatoolkit dependency in setup.py
  • PR #185 Update docs version
  • PR #194 Open source nvgraph in cugraph repository #194
  • PR #190 Added a copy option in graph creation
  • PR #196 Fix typos in readme intro
  • PR #207 mtx2csv script
  • PR #203 Added small datasets directly in the repo
  • PR #215 Simplified get_rapids_dataset_root_dir(), set a default value for the root dir
  • PR #233 Added csv datasets and edited test to use cudf for reading graphs
  • PR #247 Added some documentation for renumbering
  • PR #252 cpp test upgrades for more convenient testing on large input
  • PR #264 Add cudatoolkit conda dependency
  • PR #267 Use latest release version in update-version CI script
  • PR #270 Updated the README.md and CONTRIBUTING.md files
  • PR #281 Updated README with algorithm list

Bug Fixes

  • PR #256 Add pip to the install, clean up conda instructions
  • PR #253 Add rmm to conda configuration
  • PR #226 Bump cudf dependencies to 0.7
  • PR #169 Disable terminal output in sssp
  • PR #191 Fix double upload bug
  • PR #181 Fixed crash/rmm free error when edge values provided
  • PR #193 Fixed segfault when egde values not provided
  • PR #190 Fixed a memory reference counting error between cudf & cugraph
  • PR #190 Fixed a language level warning (cython)
  • PR #214 Removed throw exception from dtor in TC
  • PR #211 Remove hardcoded dataset paths, replace with build var that can be overridden with an env var
  • PR #206 Updated versions in conda envs
  • PR #218 Update c_graph.pyx
  • PR #224 Update erroneous comments in overlap_wrapper.pyx, woverlap_wrapper.pyx, test_louvain.py, and spectral_clustering.pyx
  • PR #220 Fixed bugs in Nvgraph triangle counting
  • PR #232 Fixed memory leaks in managing cudf columns.
  • PR #236 Fixed issue with v0.7 nightly yml environment file. Also updated the README to remove pip
  • PR #239 Added a check to prevent a cugraph object to store two different graphs.
  • PR #244 Fixed issue with nvgraph's subgraph extraction if the first vertex in the vertex list is not incident on an edge in the extracted graph
  • PR #249 Fix oudated cuDF version in gpu/build.shi
  • PR #262 Removed networkx conda dependency for both build and runtime
  • PR #271 Removed nvgraph conda dependency
  • PR #276 Removed libgdf_cffi import from bindings
  • PR #288 Add boost as a conda dependency

cuGraph 0.6.0 (22 Mar 2019)

New Features

  • PR #73 Weighted Jaccard bindings
  • PR #41 RMAT graph bindings
  • PR #43 Louvain binings
  • PR #44 SSSP bindings
  • PR #47 BSF bindings
  • PR #53 New Repo structure
  • PR #67 RMM Integration with rmm as as submodule
  • PR #82 Spectral Clustering bindings
  • PR #82 Clustering metrics binding
  • PR #85 Helper functions on python Graph object
  • PR #106 Add gpu/build.sh file for gpuCI

Improvements

  • PR #50 Reorganize directory structure to match cuDF
  • PR #85 Deleted setup.py and setup.cfg which had been replaced
  • PR #95 Code clean up
  • PR #96 Relocated mmio.c and mmio.h (external files) to thirdparty/mmio
  • PR #97 Updated python tests to speed them up
  • PR #100 Added testing for returned vertex and edge identifiers
  • PR #105 Updated python code to follow PEP8 (fixed flake8 complaints)
  • PR #121 Cleaned up READEME file
  • PR #130 Update conda build recipes
  • PR #144 Documentation for top level functions

Bug Fixes

  • PR #48 ABI Fixes
  • PR #72 Bug fix for segfault issue getting transpose from adjacency list
  • PR #105 Bug fix for memory leaks and python test failures
  • PR #110 Bug fix for segfault calling Louvain with only edge list
  • PR #115 Fixes for changes in cudf 0.6, pick up RMM from cudf instead of thirdpary
  • PR #116 Added netscience.mtx dataset to datasets.tar.gz
  • PR #120 Bug fix for segfault calling spectral clustering with only edge list
  • PR #123 Fixed weighted Jaccard to assume the input weights are given as a cudf.Series
  • PR #152 Fix conda package version string
  • PR #160 Added additional link directory to support building on CentOS-7
  • PR #221 Moved two_hop_neighbors.cuh to src folder to prevent it being installed
  • PR #223 Fixed compiler warning in cpp/src/cugraph.cu
  • PR #284 Commented out unit test code that fails due to a cudf bug

cuGraph 0.5.0 (28 Jan 2019)