remove the use of array initialize in groupby

alphamind/aggregate.pyx

@@ -51,7 +51,7 @@ cpdef list groupby(long[:] groups):
             deref(it).second.push_back(i)
 
     for v in group_ids.values():
-        res.append(array(v, dtype=np.int64))
+        res.append(v)
 
     return res
 

alphamind/portfolio/impl.pyx

@@ -13,20 +13,22 @@ cimport cython
 @cython.boundscheck(False)
 @cython.wraparound(False)
 @cython.initializedcheck(False)
-cpdef void set_value_bool(unsigned char[:, :] mat, long long[:, :] index):
+cpdef void set_value_bool(unsigned char[:, :] mat, list index, long long[:, :] used_level):
 
-    cdef size_t length = index.shape[0]
-    cdef size_t width = index.shape[1]
+    cdef size_t length = used_level.shape[0]
+    cdef size_t width = used_level.shape[1]
     cdef size_t i
     cdef size_t j
     cdef unsigned char* mat_ptr = &mat[0, 0]
-    cdef long long* index_ptr = &index[0, 0]
+    cdef long long* used_level_ptr = &used_level[0, 0]
     cdef size_t k
+    cdef size_t l
 
     for i in range(length):
         k = i * width
         for j in range(width):
-            mat_ptr[index_ptr[k + j] * width + j] = True
+            l =  index[used_level_ptr[k + j]]
+            mat_ptr[l * width + j] = True
 
 
 @cython.boundscheck(False)

alphamind/portfolio/rankbuilder.py

@@ -21,15 +21,16 @@ def rank_build(er: np.ndarray, use_rank: int, groups: np.ndarray=None) -> np.nda
         weights = zeros((length, 1))
         if groups is not None:
             group_ids = groupby(groups)
-            masks = zeros(length, dtype=bool)
+            masks = zeros((length, 1), dtype=bool)
             for current_index in group_ids:
                 current_ordering = neg_er[current_index].argsort()
-                masks[current_index[current_ordering[:use_rank]]] = True
+                current_ordering.shape = -1, 1
+                set_value_bool(masks.view(dtype=np.uint8), current_index, current_ordering[:use_rank])
             weights[masks] = 1.
         else:
             ordering = neg_er.argsort()
             weights[ordering[:use_rank]] = 1.
-        return weights
+        return weights.reshape(er.shape)
     else:
         length = er.shape[0]
         width = er.shape[1]
@@ -41,9 +42,7 @@ def rank_build(er: np.ndarray, use_rank: int, groups: np.ndarray=None) -> np.nda
             masks = zeros((length, width), dtype=bool)
             for current_index in group_ids:
                 current_ordering = neg_er[current_index].argsort(axis=0)
-                total_index = current_index[current_ordering[:use_rank]]
-                set_value_bool(masks.view(dtype=np.uint8), total_index)
-
+                set_value_bool(masks.view(dtype=np.uint8), current_index, current_ordering[:use_rank])
             for j in range(width):
                 weights[masks[:, j], j] = 1.
         else: