SHOGUN v0.9.0
|
Template class SparsePreProc, base class for preprocessors (cf. CPreProc) that apply to CSparseFeatures.
Two new functions apply_to_sparse_feature_vector() and apply_to_sparse_feature_matrix() are defined in this interface that need to be implemented in each particular preprocessor operating on CSparseFeatures.
在文件SparsePreProc.h第33行定义。
公有成员 | |
CSparsePreProc (const char *name, const char *id) | |
virtual TSparse< ST > * | apply_to_sparse_feature_matrix (CSparseFeatures< ST > *f)=0 |
virtual TSparse< ST > * | apply_to_sparse_feature_vector (TSparse< ST > *f, int32_t &len)=0 |
virtual EFeatureClass | get_feature_class () |
return that we are simple minded features (just fixed size matrices) |
CSparsePreProc | ( | const char * | name, |
const char * | id | ||
) |
virtual TSparse<ST>* apply_to_sparse_feature_matrix | ( | CSparseFeatures< ST > * | f | ) | [pure virtual] |
apply preproc on feature matrix result in feature matrix return pointer to feature_matrix, i.e. f->get_feature_matrix();
virtual TSparse<ST>* apply_to_sparse_feature_vector | ( | TSparse< ST > * | f, |
int32_t & | len | ||
) | [pure virtual] |
apply preproc on single feature vector result in feature matrix
virtual EFeatureClass get_feature_class | ( | ) | [virtual] |
return that we are simple minded features (just fixed size matrices)
实现了CPreProc。
在文件SparsePreProc.h第55行定义。