Changed the iterator to iterate over the same size tensors in the same order independent of memory layout.
This commit is contained in:
@@ -4,16 +4,16 @@
|
||||
#include "tensor.h"
|
||||
|
||||
typedef struct _tensor_scalar_iterator {
|
||||
dtype *current;
|
||||
uint32_t length;
|
||||
tensor t;
|
||||
uint32_t *index;
|
||||
} * tensoriter_scalar;
|
||||
|
||||
tensoriter_scalar tensoriter_scalar_create(tensor t);
|
||||
void tensoriter_scalar_destroy(tensoriter_scalar it);
|
||||
|
||||
bool tensoriter_scalar_next(tensoriter_scalar it);
|
||||
dtype tensoriter_scalar_get(tensoriter_scalar it);
|
||||
void tensoriter_scalar_set(tensoriter_scalar it, dtype value);
|
||||
dtype tensoriter_scalar_get(tensoriter_scalar it, bool *success);
|
||||
bool tensoriter_scalar_set(tensoriter_scalar it, dtype value);
|
||||
|
||||
void tensoriter_scalar_map(tensoriter_scalar it, dtype (*func)(dtype));
|
||||
void tensoriter_scalar_map_add(tensoriter_scalar it, dtype scalar);
|
||||
|
||||
Reference in New Issue
Block a user