pmlpp/mlpp/lin_alg/mlpp_vector.h

496 lines
10 KiB
C++

#ifndef MLPP_VECTOR_H
#define MLPP_VECTOR_H
#include "core/math/math_defs.h"
#include "core/math/math_funcs.h"
#include "core/containers/pool_vector.h"
#include "core/containers/sort_array.h"
#include "core/containers/vector.h"
#include "core/error/error_macros.h"
#include "core/os/memory.h"
#include "core/object/reference.h"
//REMOVE
#include <vector>
class MLPPMatrix;
class MLPPVector : public Reference {
GDCLASS(MLPPVector, Reference);
public:
real_t *ptrw() {
return _data;
}
const real_t *ptr() const {
return _data;
}
_FORCE_INLINE_ void push_back(real_t p_elem) {
++_size;
_data = (real_t *)memrealloc(_data, _size * sizeof(real_t));
CRASH_COND_MSG(!_data, "Out of memory");
_data[_size - 1] = p_elem;
}
_FORCE_INLINE_ void add_mlpp_vector(const Ref<MLPPVector> &p_other) {
ERR_FAIL_COND(!p_other.is_valid());
int other_size = p_other->size();
if (other_size == 0) {
return;
}
int start_offset = _size;
_size += other_size;
_data = (real_t *)memrealloc(_data, _size * sizeof(real_t));
CRASH_COND_MSG(!_data, "Out of memory");
const real_t *other_ptr = p_other->ptr();
for (int i = 0; i < other_size; ++i) {
_data[start_offset + i] = other_ptr[i];
}
}
void remove(real_t p_index) {
ERR_FAIL_INDEX(p_index, _size);
--_size;
if (_size == 0) {
memfree(_data);
_data = NULL;
return;
}
for (int i = p_index; i < _size; i++) {
_data[i] = _data[i + 1];
}
_data = (real_t *)memrealloc(_data, _size * sizeof(real_t));
CRASH_COND_MSG(!_data, "Out of memory");
}
// Removes the item copying the last value into the position of the one to
// remove. It's generally faster than `remove`.
void remove_unordered(int p_index) {
ERR_FAIL_INDEX(p_index, _size);
_size--;
if (_size == 0) {
memfree(_data);
_data = NULL;
return;
}
if (_size > p_index) {
_data[p_index] = _data[_size];
}
_data = (real_t *)memrealloc(_data, _size * sizeof(real_t));
CRASH_COND_MSG(!_data, "Out of memory");
}
void erase(const real_t &p_val) {
int idx = find(p_val);
if (idx >= 0) {
remove(idx);
}
}
int erase_multiple_unordered(const real_t &p_val) {
int from = 0;
int count = 0;
while (true) {
int64_t idx = find(p_val, from);
if (idx == -1) {
break;
}
remove_unordered(idx);
from = idx;
count++;
}
return count;
}
void invert() {
for (int i = 0; i < _size / 2; i++) {
SWAP(_data[i], _data[_size - i - 1]);
}
}
_FORCE_INLINE_ void clear() { resize(0); }
_FORCE_INLINE_ void reset() {
if (_data) {
memfree(_data);
_data = NULL;
_size = 0;
}
}
_FORCE_INLINE_ bool empty() const { return _size == 0; }
_FORCE_INLINE_ int size() const { return _size; }
void resize(int p_size) {
_size = p_size;
if (_size == 0) {
memfree(_data);
_data = NULL;
return;
}
_data = (real_t *)memrealloc(_data, _size * sizeof(real_t));
CRASH_COND_MSG(!_data, "Out of memory");
}
_FORCE_INLINE_ const real_t &operator[](int p_index) const {
CRASH_BAD_INDEX(p_index, _size);
return _data[p_index];
}
_FORCE_INLINE_ real_t &operator[](int p_index) {
CRASH_BAD_INDEX(p_index, _size);
return _data[p_index];
}
_FORCE_INLINE_ real_t get_element(int p_index) const {
ERR_FAIL_INDEX_V(p_index, _size, 0);
return _data[p_index];
}
_FORCE_INLINE_ void set_element(int p_index, real_t p_val) {
ERR_FAIL_INDEX(p_index, _size);
_data[p_index] = p_val;
}
void fill(real_t p_val) {
for (int i = 0; i < _size; i++) {
_data[i] = p_val;
}
}
void insert(int p_pos, real_t p_val) {
ERR_FAIL_INDEX(p_pos, _size + 1);
if (p_pos == _size) {
push_back(p_val);
} else {
resize(_size + 1);
for (int i = _size - 1; i > p_pos; i--) {
_data[i] = _data[i - 1];
}
_data[p_pos] = p_val;
}
}
int find(const real_t &p_val, int p_from = 0) const {
for (int i = p_from; i < _size; i++) {
if (_data[i] == p_val) {
return i;
}
}
return -1;
}
template <class C>
void sort_custom() {
int len = _size;
if (len == 0) {
return;
}
SortArray<real_t, C> sorter;
sorter.sort(_data, len);
}
void sort() {
sort_custom<_DefaultComparator<real_t>>();
}
void ordered_insert(real_t p_val) {
int i;
for (i = 0; i < _size; i++) {
if (p_val < _data[i]) {
break;
}
}
insert(i, p_val);
}
Vector<real_t> to_vector() const {
Vector<real_t> ret;
ret.resize(size());
real_t *w = ret.ptrw();
memcpy(w, _data, sizeof(real_t) * _size);
return ret;
}
PoolRealArray to_pool_vector() const {
PoolRealArray pl;
if (size()) {
pl.resize(size());
typename PoolRealArray::Write w = pl.write();
real_t *dest = w.ptr();
for (int i = 0; i < size(); ++i) {
dest[i] = static_cast<real_t>(_data[i]);
}
}
return pl;
}
Vector<uint8_t> to_byte_array() const {
Vector<uint8_t> ret;
ret.resize(_size * sizeof(real_t));
uint8_t *w = ret.ptrw();
memcpy(w, _data, sizeof(real_t) * _size);
return ret;
}
Ref<MLPPVector> duplicate() const {
Ref<MLPPVector> ret;
ret.instance();
ret->set_from_mlpp_vectorr(*this);
return ret;
}
_FORCE_INLINE_ void set_from_mlpp_vectorr(const MLPPVector &p_from) {
if (_size != p_from.size()) {
resize(p_from.size());
}
for (int i = 0; i < p_from._size; i++) {
_data[i] = p_from._data[i];
}
}
_FORCE_INLINE_ void set_from_mlpp_vector(const Ref<MLPPVector> &p_from) {
ERR_FAIL_COND(!p_from.is_valid());
if (_size != p_from->size()) {
resize(p_from->size());
}
for (int i = 0; i < p_from->_size; i++) {
_data[i] = p_from->_data[i];
}
}
_FORCE_INLINE_ void set_from_vector(const Vector<real_t> &p_from) {
if (_size != p_from.size()) {
resize(p_from.size());
}
resize(p_from.size());
for (int i = 0; i < _size; i++) {
_data[i] = p_from[i];
}
}
_FORCE_INLINE_ void set_from_pool_vector(const PoolRealArray &p_from) {
if (_size != p_from.size()) {
resize(p_from.size());
}
PoolRealArray::Read r = p_from.read();
for (int i = 0; i < _size; i++) {
_data[i] = r[i];
}
}
_FORCE_INLINE_ bool is_equal_approx(const Ref<MLPPVector> &p_with, real_t tolerance = static_cast<real_t>(CMP_EPSILON)) const {
ERR_FAIL_COND_V(!p_with.is_valid(), false);
if (unlikely(this == p_with.ptr())) {
return true;
}
if (_size != p_with->size()) {
return false;
}
for (int i = 0; i < _size; ++i) {
if (!Math::is_equal_approx(_data[i], p_with->_data[i], tolerance)) {
return false;
}
}
return true;
}
void flatten_vectors(const Vector<Ref<MLPPVector>> &A);
Ref<MLPPVector> flatten_vectorsn(const Vector<Ref<MLPPVector>> &A) const;
void hadamard_product(const Ref<MLPPVector> &b);
Ref<MLPPVector> hadamard_productn(const Ref<MLPPVector> &b) const;
void hadamard_productb(const Ref<MLPPVector> &a, const Ref<MLPPVector> &b);
void element_wise_division(const Ref<MLPPVector> &b);
Ref<MLPPVector> element_wise_divisionn(const Ref<MLPPVector> &b) const;
void element_wise_divisionb(const Ref<MLPPVector> &a, const Ref<MLPPVector> &b);
void scalar_multiply(real_t scalar);
Ref<MLPPVector> scalar_multiplyn(real_t scalar) const;
void scalar_multiplyb(real_t scalar, const Ref<MLPPVector> &a);
void scalar_add(real_t scalar);
Ref<MLPPVector> scalar_addn(real_t scalar) const;
void scalar_addb(real_t scalar, const Ref<MLPPVector> &a);
void add(const Ref<MLPPVector> &b);
Ref<MLPPVector> addn(const Ref<MLPPVector> &b) const;
void addb(const Ref<MLPPVector> &a, const Ref<MLPPVector> &b);
void sub(const Ref<MLPPVector> &b);
Ref<MLPPVector> subn(const Ref<MLPPVector> &b) const;
void subb(const Ref<MLPPVector> &a, const Ref<MLPPVector> &b);
void log();
Ref<MLPPVector> logn() const;
void logb(const Ref<MLPPVector> &a);
void log10();
Ref<MLPPVector> log10n() const;
void log10b(const Ref<MLPPVector> &a);
void exp();
Ref<MLPPVector> expn() const;
void expb(const Ref<MLPPVector> &a);
void erf();
Ref<MLPPVector> erfn() const;
void erfb(const Ref<MLPPVector> &a);
void exponentiate(real_t p);
Ref<MLPPVector> exponentiaten(real_t p) const;
void exponentiateb(const Ref<MLPPVector> &a, real_t p);
void sqrt();
Ref<MLPPVector> sqrtn() const;
void sqrtb(const Ref<MLPPVector> &a);
void cbrt();
Ref<MLPPVector> cbrtn() const;
void cbrtb(const Ref<MLPPVector> &a);
real_t dot(const Ref<MLPPVector> &b) const;
//std::vector<real_t> cross(std::vector<real_t> a, std::vector<real_t> b);
void abs();
Ref<MLPPVector> absn() const;
void absb(const Ref<MLPPVector> &a);
Ref<MLPPVector> zero_vec(int n) const;
Ref<MLPPVector> one_vec(int n) const;
Ref<MLPPVector> full_vec(int n, int k) const;
void sin();
Ref<MLPPVector> sinn() const;
void sinb(const Ref<MLPPVector> &a);
void cos();
Ref<MLPPVector> cosn() const;
void cosb(const Ref<MLPPVector> &a);
void maxv(const Ref<MLPPVector> &b);
Ref<MLPPVector> maxvn(const Ref<MLPPVector> &b) const;
void maxvb(const Ref<MLPPVector> &a, const Ref<MLPPVector> &b);
real_t max_element() const;
real_t min_element() const;
//std::vector<real_t> round(std::vector<real_t> a);
real_t euclidean_distance(const Ref<MLPPVector> &b) const;
real_t euclidean_distance_squared(const Ref<MLPPVector> &b) const;
/*
real_t norm_2(std::vector<real_t> a);
*/
real_t norm_sq() const;
real_t sum_elements() const;
//real_t cosineSimilarity(std::vector<real_t> a, std::vector<real_t> b);
void subtract_matrix_rows(const Ref<MLPPMatrix> &B);
Ref<MLPPVector> subtract_matrix_rowsn(const Ref<MLPPMatrix> &B) const;
void subtract_matrix_rowsb(const Ref<MLPPVector> &a, const Ref<MLPPMatrix> &B);
// This multiplies a, bT
Ref<MLPPMatrix> outer_product(const Ref<MLPPVector> &b) const;
// as_diagonal_matrix / to_diagonal_matrix
Ref<MLPPMatrix> diagnm() const;
String to_string();
_FORCE_INLINE_ MLPPVector() {
_size = 0;
_data = NULL;
}
_FORCE_INLINE_ MLPPVector(const MLPPVector &p_from) {
_size = 0;
_data = NULL;
resize(p_from.size());
for (int i = 0; i < p_from._size; i++) {
_data[i] = p_from._data[i];
}
}
MLPPVector(const Vector<real_t> &p_from) {
_size = 0;
_data = NULL;
resize(p_from.size());
for (int i = 0; i < _size; i++) {
_data[i] = p_from[i];
}
}
MLPPVector(const PoolRealArray &p_from) {
_size = 0;
_data = NULL;
resize(p_from.size());
typename PoolRealArray::Read r = p_from.read();
for (int i = 0; i < _size; i++) {
_data[i] = r[i];
}
}
_FORCE_INLINE_ ~MLPPVector() {
if (_data) {
reset();
}
}
// TODO: These are temporary
std::vector<real_t> to_std_vector() const;
void set_from_std_vector(const std::vector<real_t> &p_from);
MLPPVector(const std::vector<real_t> &p_from);
protected:
static void _bind_methods();
protected:
int _size;
real_t *_data;
};
#endif