Updating python to use pylibraft (rapidsai#4887)

This is a fairly sizable PR but it needs to be done before the `22.10` release as `pyraft` will not longer be released. Authors: - Corey J. Nolet (https://github.com/cjnolet) Approvers: - Dante Gama Dessavre (https://github.com/dantegd) - Mark Sadang (https://github.com/msadang) URL: rapidsai#4887
jakirkham · Sep 21, 2022 · 50e9cad · 50e9cad
1 parent 5a9ac3b
commit 50e9cad
Show file tree

Hide file tree

Showing 94 changed files with 187 additions and 176 deletions.
diff --git a/ci/gpu/build.sh b/ci/gpu/build.sh
@@ -61,7 +61,8 @@ gpuci_mamba_retry install -c conda-forge -c rapidsai -c rapidsai-nightly -c nvid
       "libraft-headers=${MINOR_VERSION}" \
       "libraft-distance=${MINOR_VERSION}" \
       "libraft-nn=${MINOR_VERSION}" \
-      "pyraft=${MINOR_VERSION}" \
+      "pylibraft=${MINOR_VERSION}" \
+      "raft-dask=${MINOR_VERSION}" \
       "dask-cudf=${MINOR_VERSION}" \
       "dask-cuda=${MINOR_VERSION}" \
       "ucx-py=${UCX_PY_VERSION}" \
@@ -126,6 +127,7 @@ if [[ -z "$PROJECT_FLASH" || "$PROJECT_FLASH" == "0" ]]; then
     set -x
     pip install "git+https://github.com/dask/distributed.git@2022.7.1" --upgrade --no-deps
     pip install "git+https://github.com/dask/dask.git@2022.7.1" --upgrade --no-deps
+    pip install "git+https://github.com/hdbscan/hdbscan.git@master" --force-reinstall --upgrade --no-deps
     set +x
 
     gpuci_logger "Python pytest for cuml"
@@ -200,6 +202,7 @@ else
     pip install "git+https://github.com/dask/distributed.git@2022.7.1" --upgrade --no-deps
     pip install "git+https://github.com/dask/dask.git@2022.7.1" --upgrade --no-deps
     pip install "git+https://github.com/dask/dask-glm@main" --force-reinstall --no-deps
+    pip install "git+https://github.com/scikit-learn-contrib/hdbscan.git@master" --force-reinstall --upgrade --no-deps
     pip install sparse
 
     set +x

diff --git a/conda/environments/cuml_dev_cuda11.0.yml b/conda/environments/cuml_dev_cuda11.0.yml
@@ -17,7 +17,8 @@ dependencies:
 - libraft-headers=22.10.*
 - libraft-distance=22.10.*
 - libraft-nn=22.10.*
-- pyraft=22.10.*
+- raft-dask=22.10.*
+- pylibraft=22.10.*
 - dask-cudf=22.10.*
 - dask-cuda=22.10.*
 - ucx>=1.13.0
@@ -33,15 +34,16 @@ dependencies:
 - treelite=2.4.0
 - statsmodels
 - seaborn
-- hdbscan
 - nltk
 - pip
 - pip:
     - git+https://github.com/dask/dask.git@2022.7.1
     - git+https://github.com/dask/distributed.git@2022.7.1
     - git+https://github.com/dask/dask-glm@main
+    - git+https://github.com/hdbscan/hdbscan.git@master
     - sparse
 
+
 # rapids-build-env, notebook-env and doc-env are defined in
 # https://docs.rapids.ai/maintainers/depmgmt/
 

diff --git a/conda/environments/cuml_dev_cuda11.2.yml b/conda/environments/cuml_dev_cuda11.2.yml
@@ -17,7 +17,8 @@ dependencies:
 - libraft-headers=22.10.*
 - libraft-distance=22.10.*
 - libraft-nn=22.10.*
-- pyraft=22.10.*
+- raft-dask=22.10.*
+- pylibraft=22.10.*
 - dask-cudf=22.10.*
 - dask-cuda=22.10.*
 - ucx>=1.13.0
@@ -33,13 +34,13 @@ dependencies:
 - treelite=2.4.0
 - statsmodels
 - seaborn
-- hdbscan
 - nltk
 - pip
 - pip:
     - git+https://github.com/dask/dask.git@2022.7.1
     - git+https://github.com/dask/distributed.git@2022.7.1
     - git+https://github.com/dask/dask-glm@main
+    - git+https://github.com/hdbscan/hdbscan.git@master
     - sparse
 
 # rapids-build-env, notebook-env and doc-env are defined in

diff --git a/conda/environments/cuml_dev_cuda11.4.yml b/conda/environments/cuml_dev_cuda11.4.yml
@@ -17,7 +17,8 @@ dependencies:
 - libraft-headers=22.10.*
 - libraft-distance=22.10.*
 - libraft-nn=22.10.*
-- pyraft=22.10.*
+- raft-dask=22.10.*
+- pylibraft=22.10.*
 - dask-cudf=22.10.*
 - dask-cuda=22.10.*
 - ucx>=1.13.0
@@ -33,13 +34,13 @@ dependencies:
 - treelite=2.4.0
 - statsmodels
 - seaborn
-- hdbscan
 - nltk
 - pip
 - pip:
     - git+https://github.com/dask/dask.git@2022.7.1
     - git+https://github.com/dask/distributed.git@2022.7.1
     - git+https://github.com/dask/dask-glm.git@main
+    - git+https://github.com/hdbscan/hdbscan.git@master
     - sparse
 
 # rapids-build-env, notebook-env and doc-env are defined in

diff --git a/conda/environments/cuml_dev_cuda11.5.yml b/conda/environments/cuml_dev_cuda11.5.yml
@@ -17,7 +17,8 @@ dependencies:
 - libraft-headers=22.10.*
 - libraft-distance=22.10.*
 - libraft-nn=22.10.*
-- pyraft=22.10.*
+- raft-dask=22.10.*
+- pylibraft=22.10.*
 - dask-cudf=22.10.*
 - dask-cuda=22.10.*
 - ucx>=1.13.0
@@ -33,13 +34,13 @@ dependencies:
 - treelite=2.4.0
 - statsmodels
 - seaborn
-- hdbscan
 - nltk
 - pip
 - pip:
     - git+https://github.com/dask/dask.git@2022.7.1
     - git+https://github.com/dask/distributed.git@2022.7.1
     - git+https://github.com/dask/dask-glm@main
+    - git+https://github.com/hdbscan/hdbscan.git@master
     - sparse
 
 # rapids-build-env, notebook-env and doc-env are defined in

diff --git a/conda/recipes/cuml/meta.yaml b/conda/recipes/cuml/meta.yaml
@@ -40,7 +40,8 @@ requirements:
     - cudf {{ minor_version }}
     - libcuml={{ version }}
     - libcumlprims {{ minor_version }}
-    - pyraft {{ minor_version }}
+    - pylibraft {{ minor_version }}
+    - raft-dask {{ minor_version }}
     - cudatoolkit {{ cuda_version }}.*
     - ucx-py {{ ucx_py_version }}
     - ucx-proc=*=gpu
@@ -51,7 +52,8 @@ requirements:
     - dask-cudf {{ minor_version }}
     - libcuml={{ version }}
     - libcumlprims {{ minor_version }}
-    - pyraft {{ minor_version }}
+    - pylibraft {{ minor_version }}
+    - raft-dask {{ minor_version }}
     - cupy>=7.8.0,<12.0.0a0
     - treelite=2.4.0
     - nccl>=2.9.9

diff --git a/cpp/src/hdbscan/detail/reachability.cuh b/cpp/src/hdbscan/detail/reachability.cuh
@@ -24,7 +24,6 @@
 #include <raft/linalg/unary_op.cuh>
 
 #include <raft/sparse/convert/csr.cuh>
-#include <raft/sparse/hierarchy/detail/connectivities.cuh>
 #include <raft/sparse/linalg/symmetrize.cuh>
 #include <raft/sparse/selection/knn_graph.cuh>
 

diff --git a/cpp/src/hdbscan/runner.h b/cpp/src/hdbscan/runner.h
@@ -23,9 +23,9 @@
 
 #include <cuml/common/logger.hpp>
 
+#include <raft/cluster/detail/agglomerative.cuh>
+#include <raft/cluster/detail/mst.cuh>
 #include <raft/sparse/coo.hpp>
-#include <raft/sparse/hierarchy/detail/agglomerative.cuh>
-#include <raft/sparse/hierarchy/detail/mst.cuh>
 
 #include "detail/condense.cuh"
 #include "detail/extract.cuh"
@@ -160,35 +160,35 @@ void build_linkage(const raft::handle_t& handle,
   rmm::device_uvector<value_idx> color(m, stream);
   FixConnectivitiesRedOp<value_idx, value_t> red_op(color.data(), core_dists, m);
   // during knn graph connection
-  raft::hierarchy::detail::build_sorted_mst(handle,
-                                            X,
-                                            mutual_reachability_indptr.data(),
-                                            mutual_reachability_coo.cols(),
-                                            mutual_reachability_coo.vals(),
-                                            m,
-                                            n,
-                                            out.get_mst_src(),
-                                            out.get_mst_dst(),
-                                            out.get_mst_weights(),
-                                            color.data(),
-                                            mutual_reachability_coo.nnz,
-                                            red_op,
-                                            metric,
-                                            (size_t)10);
+  raft::cluster::detail::build_sorted_mst(handle,
+                                          X,
+                                          mutual_reachability_indptr.data(),
+                                          mutual_reachability_coo.cols(),
+                                          mutual_reachability_coo.vals(),
+                                          m,
+                                          n,
+                                          out.get_mst_src(),
+                                          out.get_mst_dst(),
+                                          out.get_mst_weights(),
+                                          color.data(),
+                                          mutual_reachability_coo.nnz,
+                                          red_op,
+                                          metric,
+                                          (size_t)10);
 
   /**
    * Perform hierarchical labeling
    */
   size_t n_edges = m - 1;
 
-  raft::hierarchy::detail::build_dendrogram_host(handle,
-                                                 out.get_mst_src(),
-                                                 out.get_mst_dst(),
-                                                 out.get_mst_weights(),
-                                                 n_edges,
-                                                 out.get_children(),
-                                                 out.get_deltas(),
-                                                 out.get_sizes());
+  raft::cluster::detail::build_dendrogram_host(handle,
+                                               out.get_mst_src(),
+                                               out.get_mst_dst(),
+                                               out.get_mst_weights(),
+                                               n_edges,
+                                               out.get_children(),
+                                               out.get_deltas(),
+                                               out.get_sizes());
 }
 
 template <typename value_idx = int64_t, typename value_t = float>

diff --git a/cpp/src/hierarchy/linkage.cu b/cpp/src/hierarchy/linkage.cu
@@ -16,7 +16,7 @@
 
 #include "pw_dist_graph.cuh"
 #include <cuml/cluster/linkage.hpp>
-#include <raft/sparse/hierarchy/single_linkage.cuh>
+#include <raft/cluster/single_linkage.cuh>
 
 namespace raft {
 class handle_t;
@@ -28,34 +28,34 @@ void single_linkage_pairwise(const raft::handle_t& handle,
                              const float* X,
                              size_t m,
                              size_t n,
-                             raft::hierarchy::linkage_output<int, float>* out,
+                             raft::cluster::linkage_output<int, float>* out,
                              raft::distance::DistanceType metric,
                              int n_clusters)
 {
-  raft::hierarchy::single_linkage<int, float, raft::hierarchy::LinkageDistance::PAIRWISE>(
+  raft::cluster::single_linkage<int, float, raft::cluster::LinkageDistance::PAIRWISE>(
     handle, X, m, n, metric, out, 0, n_clusters);
 }
 
 void single_linkage_neighbors(const raft::handle_t& handle,
                               const float* X,
                               size_t m,
                               size_t n,
-                              raft::hierarchy::linkage_output<int, float>* out,
+                              raft::cluster::linkage_output<int, float>* out,
                               raft::distance::DistanceType metric,
                               int c,
                               int n_clusters)
 {
-  raft::hierarchy::single_linkage<int, float, raft::hierarchy::LinkageDistance::KNN_GRAPH>(
+  raft::cluster::single_linkage<int, float, raft::cluster::LinkageDistance::KNN_GRAPH>(
     handle, X, m, n, metric, out, c, n_clusters);
 }
 
 struct distance_graph_impl_int_float
-  : public raft::hierarchy::detail::
-      distance_graph_impl<raft::hierarchy::LinkageDistance::PAIRWISE, int, float> {
+  : public raft::cluster::detail::
+      distance_graph_impl<raft::cluster::LinkageDistance::PAIRWISE, int, float> {
 };
 struct distance_graph_impl_int_double
-  : public raft::hierarchy::detail::
-      distance_graph_impl<raft::hierarchy::LinkageDistance::PAIRWISE, int, double> {
+  : public raft::cluster::detail::
+      distance_graph_impl<raft::cluster::LinkageDistance::PAIRWISE, int, double> {
 };
 
 };  // end namespace ML
diff --git a/cpp/src/hierarchy/pw_dist_graph.cuh b/cpp/src/hierarchy/pw_dist_graph.cuh
@@ -28,8 +28,8 @@
 
 // TODO: Not a good strategy for pluggability but will be
 // removed once our dense pairwise distance API is in RAFT
-#include <raft/sparse/hierarchy/common.h>
-#include <raft/sparse/hierarchy/detail/connectivities.cuh>
+#include <raft/cluster/detail/connectivities.cuh>
+#include <raft/cluster/single_linkage_types.hpp>
 
 #include <thrust/device_ptr.h>
 #include <thrust/execution_policy.h>
@@ -44,7 +44,7 @@
 #include <limits>
 
 namespace raft {
-namespace hierarchy {
+namespace cluster {
 namespace detail {
 
 template <typename value_idx>
@@ -117,7 +117,7 @@ void pairwise_distances(const raft::handle_t& handle,
  * @tparam value_t
  */
 template <typename value_idx, typename value_t>
-struct distance_graph_impl<raft::hierarchy::LinkageDistance::PAIRWISE, value_idx, value_t> {
+struct distance_graph_impl<raft::cluster::LinkageDistance::PAIRWISE, value_idx, value_t> {
   void run(const raft::handle_t& handle,
            const value_t* X,
            size_t m,
@@ -140,5 +140,5 @@ struct distance_graph_impl<raft::hierarchy::LinkageDistance::PAIRWISE, value_idx
 };
 
 };  // namespace detail
-};  // end namespace hierarchy
+};  // namespace cluster
 };  // end namespace raft
diff --git a/cpp/test/sg/hdbscan_test.cu b/cpp/test/sg/hdbscan_test.cu
@@ -31,7 +31,7 @@
 
 #include <metrics/adjusted_rand_index.cuh>
 
-#include <raft/sparse/hierarchy/detail/agglomerative.cuh>
+#include <raft/cluster/detail/agglomerative.cuh>
 
 #include <raft/distance/distance_type.hpp>
 #include <raft/linalg/transpose.cuh>
@@ -177,14 +177,14 @@ class ClusterCondensingTest : public ::testing::TestWithParam<ClusterCondensingI
     /**
      * Build dendrogram of MST
      */
-    raft::hierarchy::detail::build_dendrogram_host(handle,
-                                                   mst_src.data(),
-                                                   mst_dst.data(),
-                                                   mst_data.data(),
-                                                   params.n_row - 1,
-                                                   out_children.data(),
-                                                   out_delta.data(),
-                                                   out_size.data());
+    raft::cluster::detail::build_dendrogram_host(handle,
+                                                 mst_src.data(),
+                                                 mst_dst.data(),
+                                                 mst_data.data(),
+                                                 params.n_row - 1,
+                                                 out_children.data(),
+                                                 out_delta.data(),
+                                                 out_size.data());
 
     /**
      * Condense Hierarchy

diff --git a/python/cuml/cluster/agglomerative.pyx b/python/cuml/cluster/agglomerative.pyx
@@ -23,7 +23,7 @@ import numpy as np
 from cuml.common.array import CumlArray
 from cuml.common.base import Base
 from cuml.common.doc_utils import generate_docstring
-from raft.common.handle cimport handle_t
+from pylibraft.common.handle cimport handle_t
 from cuml.common import input_to_cuml_array
 from cuml.common.array_descriptor import CumlArrayDescriptor
 from cuml.common.mixins import ClusterMixin

diff --git a/python/cuml/cluster/dbscan.pyx b/python/cuml/cluster/dbscan.pyx
@@ -27,7 +27,7 @@ from libc.stdlib cimport calloc, malloc, free
 from cuml.common.array import CumlArray
 from cuml.common.base import Base
 from cuml.common.doc_utils import generate_docstring
-from raft.common.handle cimport handle_t
+from pylibraft.common.handle cimport handle_t
 from cuml.common import input_to_cuml_array
 from cuml.common import using_output_type
 from cuml.common.array_descriptor import CumlArrayDescriptor

diff --git a/python/cuml/cluster/hdbscan.pyx b/python/cuml/cluster/hdbscan.pyx
@@ -26,9 +26,9 @@ import cupy as cp
 from cuml.common.array import CumlArray
 from cuml.common.base import Base
 from cuml.common.doc_utils import generate_docstring
-from raft.common.handle cimport handle_t
+from pylibraft.common.handle cimport handle_t
 
-from raft.common.handle import Handle
+from pylibraft.common.handle import Handle
 from cuml.common import input_to_cuml_array
 from cuml.common.array_descriptor import CumlArrayDescriptor
 from cuml.common.mixins import ClusterMixin

diff --git a/python/cuml/cluster/kmeans.pyx b/python/cuml/cluster/kmeans.pyx
@@ -34,7 +34,7 @@ from cuml.common.mixins import ClusterMixin
 from cuml.common.mixins import CMajorInputTagMixin
 from cuml.common import input_to_cuml_array
 from cuml.cluster.kmeans_utils cimport *
-from raft.common.handle cimport handle_t
+from pylibraft.common.handle cimport handle_t
 
 cdef extern from "cuml/cluster/kmeans.hpp" namespace "ML::kmeans":
 

diff --git a/python/cuml/cluster/kmeans_mg.pyx b/python/cuml/cluster/kmeans_mg.pyx
@@ -28,7 +28,7 @@ from libc.stdlib cimport calloc, malloc, free
 
 from cuml.common.array import CumlArray
 from cuml.common.base import Base
-from raft.common.handle cimport handle_t
+from pylibraft.common.handle cimport handle_t
 from cuml.common import input_to_cuml_array
 
 from cuml.cluster import KMeans