我在对我的代码进行 cythonizing 时遇到问题,更具体地说是以下(和类似的)代码片段:
cdef double [:,:] grad_d_him_d_jm
grad_d_ihm_d_jm = grad_d_im_d_jm(...)
其中 grad_d_im_d_jm(...) 将返回一个双 [:,:] 内存视图。此代码将由 Cython 翻译成以下 C 代码:
__pyx_t_1 = __pyx_f_24gradient_better_c_mviews_grad_d_im_d_jm(__pyx_v_i, __pyx_v_j, __pyx_v_m, __pyx_v_structure, __pyx_v_distances);
if (unlikely(!__pyx_t_1)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 203; __pyx_clineno = __LINE__; goto __pyx_L1_error;}
__Pyx_GOTREF(__pyx_t_1);
__pyx_t_7 = __Pyx_PyObject_to_MemoryviewSlice_dsds_double(__pyx_t_1);
if (unlikely(!__pyx_t_7.memview)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 203; __pyx_clineno = __LINE__; goto __pyx_L1_error;}
__Pyx_DECREF(__pyx_t_1); __pyx_t_1 = 0;
__pyx_v_grad_d_ihm_d_jm = __pyx_t_7;
__pyx_t_7.memview = NULL;
__pyx_t_7.data = NULL;
当我在循环中执行此操作时,我怀疑 Python API 调用会对我的代码速度产生相当大的影响。
GOTREF / DECREF 调用也发生在其他场合,以及 PyFloat_asFloat:
cdef float sp
sp = scalar_product()
其中 scalar_product() 返回一个 cdef 浮点数。这个片段被翻译成
__pyx_t_1 = __pyx_f_24gradient_better_c_mviews_scalar_product(__pyx_v_i, __pyx_v_j, __pyx_v_m, __pyx_v_structure);
if (unlikely(!__pyx_t_1)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 178; __pyx_clineno = __LINE__; goto __pyx_L1_error;}
__Pyx_GOTREF(__pyx_t_1);
__pyx_t_2 = __pyx_PyFloat_AsFloat(__pyx_t_1);
if (unlikely((__pyx_t_2 == (float)-1) && PyErr_Occurred())) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 178; __pyx_clineno = __LINE__; goto __pyx_L1_error;}
__Pyx_DECREF(__pyx_t_1); __pyx_t_1 = 0;
__pyx_v_sp = __pyx_t_2;
我正在运行 Python 2.7.11+ 和 Cython 0.23.4。如果您能告诉我 a) 这与性能无关或 b) 如何解决它,我将不胜感激。如果我能改进这个问题,请告诉我,我很乐意这样做。