[MAINT] Update Ctypes in line with scikit-learn: main (#141)

* Update submodule and c-types in tree/

---------

Signed-off-by: Adam Li <adam2392@gmail.com>

.github/workflows/build_wheels.yml

@@ -44,7 +44,7 @@ jobs:
           # - [macos-12, macosx_*, arm64]
           - [windows-2019, win, AMD64]
 
-        python: [["cp39", "3.9"], ["cp310", "3.10"], ["cp311", "3.11.0-alpha - 3.11.0"]]
+        python: [["cp39", "3.9"], ["cp310", "3.10"], ["cp311", "3.11"]]
         # python[0] is used to specify the python versions made by cibuildwheel
 
     env:

.github/workflows/main.yml

@@ -100,7 +100,7 @@ jobs:
 
       - name: Setup build and install scikit-tree
         run: |
-          ./spin build -j 2 --forcesubmodule
+          ./spin build -j 2
 
       - name: Ccache performance
         shell: bash -l {0}

.spin/cmds.py

@@ -54,7 +54,7 @@ def setup_submodule(forcesubmodule=False):
     This will update the submodule, which then must be commited so that
     git knows the submodule needs to be at a certain commit hash.
     """
-    commit_fpath = "./sktree/_lib/sklearn_fork/commit.txt"
+    commit_fpath = "./sktree/_lib/commit.txt"
     submodule = "./sktree/_lib/sklearn_fork"
     commit = ""
     current_hash = ""

build_requirements.txt

@@ -1,9 +1,9 @@
 meson
 meson-python
-cython>=3.0
+cython==0.29.36
 ninja
 numpy
-scikit-learn>=1.3
+scikit-learn>=1.3.1
 click
 rich-click
 doit

pyproject.toml

@@ -38,7 +38,7 @@ include = [
 dependencies = [
     'numpy',
     'scipy>=1.5.0',
-    'scikit-learn>=1.3'
+    'scikit-learn>=1.3.1'
 ]
 
 

requirements.txt

@@ -1,3 +1,3 @@
 numpy>=1.25
 scipy
-scikit-learn>=1.3
+scikit-learn>=1.3.1

sktree/__init__.py

@@ -59,6 +59,7 @@
         )
         from .ensemble._honest_forest import HonestForestClassifier
     except ImportError as e:
+        print(e.msg)
         msg = """Error importing scikit-tree: you cannot import scikit-tree while
         being in scikit-tree source directory; please exit the scikit-tree source
         tree first and relaunch your Python interpreter."""

sktree/_lib/meson.build

@@ -42,4 +42,36 @@ foreach py_source: python_sources
     './sklearn/ensemble/' + py_source,
     subdir: 'sktree/_lib/sklearn/ensemble'
   )
-endforeach
+endforeach
+
+# TODO: Can remove if included in scikit-learn eventually
+# install tree/ submodule
+extensions = [
+  '_quad_tree',
+]
+
+foreach ext: extensions
+  py3.extension_module(ext,
+    cython_gen_cpp.process('./sklearn/neighbors/' + ext + '.pyx'),
+    c_args: cython_c_args,
+    include_directories: [incdir_numpy,],
+    install: true,
+    subdir: 'sktree/_lib/sklearn/neighbors/',
+  )
+endforeach
+
+# install tree/ submodule
+extensions = [
+  '_typedefs',
+  '_random',
+]
+
+foreach ext: extensions
+  py3.extension_module(ext,
+    cython_gen_cpp.process('./sklearn/utils/' + ext + '.pyx'),
+    c_args: cython_c_args,
+    include_directories: [incdir_numpy,],
+    install: true,
+    subdir: 'sktree/_lib/sklearn/utils/',
+  )
+endforeach

sktree/tree/_marginal.pxd

@@ -2,19 +2,16 @@ import numpy as np
 
 cimport numpy as cnp
 
-from .._lib.sklearn.tree._tree cimport DOUBLE_t  # Type of y, sample_weight
-from .._lib.sklearn.tree._tree cimport DTYPE_t  # Type of X
-from .._lib.sklearn.tree._tree cimport INT32_t  # Signed 32 bit integer
-from .._lib.sklearn.tree._tree cimport SIZE_t  # Type for indices and counters
-from .._lib.sklearn.tree._tree cimport UINT32_t  # Unsigned 32 bit integer
 from .._lib.sklearn.tree._tree cimport BaseTree, Node
+from .._lib.sklearn.tree._utils cimport UINT32_t
+from .._lib.sklearn.utils._typedefs cimport float32_t, float64_t, intp_t
 
 
 cpdef apply_marginal_tree(
     BaseTree tree,
     object X,
-    const SIZE_t[:] marginal_indices,
-    int traversal_method,
+    const intp_t[:] marginal_indices,
+    intp_t traversal_method,
     unsigned char use_sample_weight,
     object random_state
 )

sktree/tree/_marginal.pyx

@@ -18,15 +18,15 @@ from numpy import float32 as DTYPE
 
 TREE_LEAF = -1
 TREE_UNDEFINED = -2
-cdef SIZE_t _TREE_LEAF = TREE_LEAF
-cdef SIZE_t _TREE_UNDEFINED = TREE_UNDEFINED
+cdef intp_t _TREE_LEAF = TREE_LEAF
+cdef intp_t _TREE_UNDEFINED = TREE_UNDEFINED
 
 
 cpdef apply_marginal_tree(
     BaseTree tree,
     object X,
-    const SIZE_t[:] marginal_indices,
-    int traversal_method,
+    const intp_t[:] marginal_indices,
+    intp_t traversal_method,
     unsigned char use_sample_weight,
     object random_state
 ):
@@ -41,7 +41,7 @@ cpdef apply_marginal_tree(
     marginal_indices : ndarray of shape (n_marginals,)
         The indices of the features to marginalize, which
         are columns in ``X``.
-    traversal_method : int
+    traversal_method : intp_t
         The traversal method to use. 0 for 'random', 1 for
         'weighted'.
     use_sample_weight : unsigned char
@@ -62,13 +62,13 @@ cpdef apply_marginal_tree(
     if X.dtype != DTYPE:
         raise ValueError("X.dtype should be np.float32, got %s" % X.dtype)
 
-    cdef SIZE_t n_marginals = marginal_indices.shape[0]
+    cdef intp_t n_marginals = marginal_indices.shape[0]
 
     # sklearn_rand_r random number state
     cdef UINT32_t rand_r_state = random_state.randint(0, RAND_R_MAX)
 
     # define a set of all marginal indices
-    cdef unordered_set[SIZE_t] marginal_indices_map
+    cdef unordered_set[intp_t] marginal_indices_map
 
     # check all marginal indices are valid, and also convert to an unordered map
     for i in range(n_marginals):
@@ -94,19 +94,19 @@ cpdef apply_marginal_tree(
 cdef void _resample_split_node(
     BaseTree tree,
     Node* node,
-    unordered_set[SIZE_t] marginal_indices_map,
-    const DTYPE_t[:, :] X,
-    const DOUBLE_t[:, ::1] y,
-    const DOUBLE_t[:] sample_weight,
+    unordered_set[intp_t] marginal_indices_map,
+    const float32_t[:, :] X,
+    const float64_t[:, ::1] y,
+    const float64_t[:] sample_weight,
 ) noexcept nogil:
     pass
 
 
 cdef inline cnp.ndarray _apply_dense_marginal(
     BaseTree tree,
-    const DTYPE_t[:, :] X,
-    unordered_set[SIZE_t] marginal_indices_map,
-    int traversal_method,
+    const float32_t[:, :] X,
+    unordered_set[intp_t] marginal_indices_map,
+    intp_t traversal_method,
     unsigned char use_sample_weight,
     UINT32_t* rand_r_state
 ):
@@ -122,10 +122,10 @@ cdef inline cnp.ndarray _apply_dense_marginal(
         The tree to apply.
     X : const ndarray of shape (n_samples, n_features)
         The data matrix.
-    marginal_indices_map : unordered_set[SIZE_t]
+    marginal_indices_map : unordered_set[intp_t]
         The indices of the features to marginalize, which
         are columns in ``X``.
-    traversal_method : int
+    traversal_method : intp_t
         The traversal method to use. 0 for 'random', 1 for
         'weighted'.
     use_sample_weight : unsigned char
@@ -135,20 +135,20 @@ cdef inline cnp.ndarray _apply_dense_marginal(
         The random number state.
     """
     # Extract input
-    cdef const DTYPE_t[:, :] X_ndarray = X
-    cdef SIZE_t n_samples = X.shape[0]
-    cdef DTYPE_t X_i_node_feature
+    cdef const float32_t[:, :] X_ndarray = X
+    cdef intp_t n_samples = X.shape[0]
+    cdef float32_t X_i_node_feature
 
-    cdef DTYPE_t n_node_samples, n_right_samples, n_left_samples
-    cdef double p_left
-    cdef int is_left
+    cdef float32_t n_node_samples, n_right_samples, n_left_samples
+    cdef float64_t p_left
+    cdef intp_t is_left
 
     # Initialize output
-    cdef SIZE_t[:] out = np.zeros(n_samples, dtype=np.intp)
+    cdef intp_t[:] out = np.zeros(n_samples, dtype=np.intp)
 
     # Initialize auxiliary data-structure
     cdef Node* node = NULL
-    cdef SIZE_t i = 0
+    cdef intp_t i = 0
 
     with nogil:
         for i in prange(n_samples):
@@ -172,7 +172,7 @@ cdef inline cnp.ndarray _apply_dense_marginal(
                             n_right_samples = tree.nodes[node.right_child].n_node_samples
 
                         # compute the probabilies for going left and right
-                        p_left = (<double>n_left_samples / n_node_samples)
+                        p_left = (<float64_t>n_left_samples / n_node_samples)
 
                         # randomly sample a direction
                         is_left = rand_weighted_binary(p_left, rand_r_state)
@@ -202,14 +202,14 @@ cdef inline cnp.ndarray _apply_dense_marginal(
                     else:
                         node = &tree.nodes[node.right_child]
 
-            out[i] = <SIZE_t>(node - tree.nodes)  # node offset
+            out[i] = <intp_t>(node - tree.nodes)  # node offset
 
     return np.asarray(out)
 
 
-cdef inline int is_element_present(unordered_set[SIZE_t]& my_set, SIZE_t element) noexcept nogil:
+cdef inline intp_t is_element_present(unordered_set[intp_t]& my_set, intp_t element) noexcept nogil:
     """Helper function to check presence of element in set."""
-    cdef unordered_set[SIZE_t].iterator it = my_set.find(element)
+    cdef unordered_set[intp_t].iterator it = my_set.find(element)
 
     if it != my_set.end():
         return 1