From eb85684feb0505694e66365ba9f4d10a409f8f0b Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?J=C3=A9r=C3=A9mie=20du=20Boisberranger?= <34657725+jeremiedbb@users.noreply.github.com> Date: Wed, 11 Oct 2023 11:43:47 +0200 Subject: [PATCH] MAINT Follow cython performance hints (#27521) --- .../_middle_term_computer.pyx.tp | 2 +- .../preprocessing/_csr_polynomial_expansion.pyx | 2 +- sklearn/tree/_splitter.pyx | 2 +- sklearn/tree/_utils.pxd | 2 +- sklearn/tree/_utils.pyx | 14 ++++++-------- 5 files changed, 10 insertions(+), 12 deletions(-) diff --git a/sklearn/metrics/_pairwise_distances_reduction/_middle_term_computer.pyx.tp b/sklearn/metrics/_pairwise_distances_reduction/_middle_term_computer.pyx.tp index f2d89ed65909c..bfe465394c4d2 100644 --- a/sklearn/metrics/_pairwise_distances_reduction/_middle_term_computer.pyx.tp +++ b/sklearn/metrics/_pairwise_distances_reduction/_middle_term_computer.pyx.tp @@ -82,7 +82,7 @@ cdef void _middle_term_sparse_dense_{{name_suffix}}( intp_t Y_end, bint c_ordered_middle_term, float64_t * dist_middle_terms, -) nogil: +) noexcept nogil: # This routine assumes that dist_middle_terms is a pointer to the first element # of a buffer filled with zeros of length at least equal to n_X × n_Y, conceptually # representing a 2-d C-ordered of F-ordered array. diff --git a/sklearn/preprocessing/_csr_polynomial_expansion.pyx b/sklearn/preprocessing/_csr_polynomial_expansion.pyx index 90f81c0399a6e..017af83f035b2 100644 --- a/sklearn/preprocessing/_csr_polynomial_expansion.pyx +++ b/sklearn/preprocessing/_csr_polynomial_expansion.pyx @@ -166,7 +166,7 @@ cpdef void _csr_polynomial_expansion( INDEX_B_t[:] result_indptr, # OUT FLAG_t interaction_only, FLAG_t degree -) nogil: +): """ Perform a second or third degree polynomial or interaction expansion on a compressed sparse row (CSR) matrix. The method used only takes products of diff --git a/sklearn/tree/_splitter.pyx b/sklearn/tree/_splitter.pyx index a9d3a169ec84a..81cd61ed22631 100644 --- a/sklearn/tree/_splitter.pyx +++ b/sklearn/tree/_splitter.pyx @@ -263,7 +263,7 @@ cdef inline void shift_missing_values_to_left_if_required( SplitRecord* best, intp_t[::1] samples, intp_t end, -) nogil: +) noexcept nogil: cdef intp_t i, p, current_end # The partitioner partitions the data such that the missing values are in # samples[-n_missing:] for the criterion to consume. If the missing values diff --git a/sklearn/tree/_utils.pxd b/sklearn/tree/_utils.pxd index 4167230bfbf4d..b59d18879ca94 100644 --- a/sklearn/tree/_utils.pxd +++ b/sklearn/tree/_utils.pxd @@ -39,7 +39,7 @@ ctypedef fused realloc_ptr: (Cell*) (Node**) -cdef realloc_ptr safe_realloc(realloc_ptr* p, size_t nelems) except * nogil +cdef int safe_realloc(realloc_ptr* p, size_t nelems) except -1 nogil cdef cnp.ndarray sizet_ptr_to_ndarray(intp_t* data, intp_t size) diff --git a/sklearn/tree/_utils.pyx b/sklearn/tree/_utils.pyx index 3c0c312b25fbe..b6115d4f21d18 100644 --- a/sklearn/tree/_utils.pyx +++ b/sklearn/tree/_utils.pyx @@ -22,22 +22,20 @@ from ..utils._random cimport our_rand_r # Helper functions # ============================================================================= -cdef realloc_ptr safe_realloc(realloc_ptr* p, size_t nelems) except * nogil: +cdef int safe_realloc(realloc_ptr* p, size_t nelems) except -1 nogil: # sizeof(realloc_ptr[0]) would be more like idiomatic C, but causes Cython # 0.20.1 to crash. cdef size_t nbytes = nelems * sizeof(p[0][0]) if nbytes / sizeof(p[0][0]) != nelems: # Overflow in the multiplication - with gil: - raise MemoryError("could not allocate (%d * %d) bytes" - % (nelems, sizeof(p[0][0]))) + raise MemoryError(f"could not allocate ({nelems} * {sizeof(p[0][0])}) bytes") + cdef realloc_ptr tmp = realloc(p[0], nbytes) if tmp == NULL: - with gil: - raise MemoryError("could not allocate %d bytes" % nbytes) + raise MemoryError(f"could not allocate {nbytes} bytes") p[0] = tmp - return tmp # for convenience + return 0 def _realloc_test(): @@ -111,7 +109,7 @@ cdef class WeightedPQueue: or 0 otherwise. """ self.array_ptr = 0 - # Since safe_realloc can raise MemoryError, use `except *` + # Since safe_realloc can raise MemoryError, use `except -1` safe_realloc(&self.array_, self.capacity) return 0