#ifndef XGBOOST_COMMON_CATEGORICAL_H_
#define XGBOOST_COMMON_CATEGORICAL_H_
#include "bitfield.h"
#include "xgboost/base.h"
#include "xgboost/data.h"
#include "xgboost/span.h"
#include "xgboost/tree_model.h"
namespace xgboost::common {
using CatBitField = LBitField32;
using KCatBitField = CLBitField32;
template <typename T>
XGBOOST_DEVICE bst_cat_t AsCat(T const& v) {
return static_cast<bst_cat_t>(v);
}
inline XGBOOST_DEVICE bool IsCat(Span<FeatureType const> ft, bst_feature_t fidx) {
return !ft.empty() && ft[fidx] == FeatureType::kCategorical;
}
constexpr inline bst_cat_t OutOfRangeCat() {
return static_cast<bst_cat_t>(16777217) - static_cast<bst_cat_t>(1);
}
inline XGBOOST_DEVICE bool InvalidCat(float cat) {
constexpr auto kMaxCat = OutOfRangeCat();
static_assert(static_cast<bst_cat_t>(static_cast<float>(kMaxCat)) == kMaxCat);
static_assert(static_cast<bst_cat_t>(static_cast<float>(kMaxCat + 1)) != kMaxCat + 1);
static_assert(static_cast<float>(kMaxCat + 1) == kMaxCat);
return cat < 0 || cat >= kMaxCat;
}
inline XGBOOST_DEVICE bool Decision(common::Span<CatBitField::value_type const> cats, float cat) {
KCatBitField const s_cats(cats);
if (XGBOOST_EXPECT(InvalidCat(cat), false)) {
return true;
}
auto pos = KCatBitField::ToBitPos(cat);
if (pos.int_pos >= cats.size()) {
return true;
}
return !s_cats.Check(AsCat(cat));
}
inline void InvalidCategory() {
auto str = std::to_string(OutOfRangeCat());
LOG(FATAL) << "Invalid categorical value detected. Categorical value should be non-negative, "
"less than total number of categories in training data and less than " +
str;
}
inline void CheckMaxCat(float max_cat, size_t n_categories) {
CHECK_GE(max_cat + 1, n_categories)
<< "Maximum cateogry should not be lesser than the total number of categories.";
}
XGBOOST_DEVICE inline bool UseOneHot(uint32_t n_cats, uint32_t max_cat_to_onehot) {
bool use_one_hot = n_cats < max_cat_to_onehot;
return use_one_hot;
}
struct IsCatOp {
XGBOOST_DEVICE bool operator()(FeatureType ft) { return ft == FeatureType::kCategorical; }
};
inline auto GetNodeCats(common::Span<CatBitField::value_type const> categories,
RegTree::CategoricalSplitMatrix::Segment seg) {
KCatBitField node_cats{categories.subspan(seg.beg, seg.size)};
return node_cats;
}
}
#endif