2023-12-30 00:41:59 +01:00
|
|
|
/*************************************************************************/
|
|
|
|
/* outlier_finder.cpp */
|
|
|
|
/*************************************************************************/
|
|
|
|
/* This file is part of: */
|
|
|
|
/* PMLPP Machine Learning Library */
|
|
|
|
/* https://github.com/Relintai/pmlpp */
|
|
|
|
/*************************************************************************/
|
2023-12-30 00:43:39 +01:00
|
|
|
/* Copyright (c) 2023-present Péter Magyar. */
|
2023-12-30 00:41:59 +01:00
|
|
|
/* Copyright (c) 2022-2023 Marc Melikyan */
|
|
|
|
/* */
|
|
|
|
/* Permission is hereby granted, free of charge, to any person obtaining */
|
|
|
|
/* a copy of this software and associated documentation files (the */
|
|
|
|
/* "Software"), to deal in the Software without restriction, including */
|
|
|
|
/* without limitation the rights to use, copy, modify, merge, publish, */
|
|
|
|
/* distribute, sublicense, and/or sell copies of the Software, and to */
|
|
|
|
/* permit persons to whom the Software is furnished to do so, subject to */
|
|
|
|
/* the following conditions: */
|
|
|
|
/* */
|
|
|
|
/* The above copyright notice and this permission notice shall be */
|
|
|
|
/* included in all copies or substantial portions of the Software. */
|
|
|
|
/* */
|
|
|
|
/* THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, */
|
|
|
|
/* EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF */
|
|
|
|
/* MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.*/
|
|
|
|
/* IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY */
|
|
|
|
/* CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, */
|
|
|
|
/* TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE */
|
|
|
|
/* SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE. */
|
|
|
|
/*************************************************************************/
|
2023-01-23 21:13:26 +01:00
|
|
|
|
2023-01-24 18:12:23 +01:00
|
|
|
#include "outlier_finder.h"
|
2023-02-09 15:30:33 +01:00
|
|
|
|
2023-01-24 18:12:23 +01:00
|
|
|
#include "../stat/stat.h"
|
2023-01-23 21:13:26 +01:00
|
|
|
|
2023-02-09 15:30:33 +01:00
|
|
|
real_t MLPPOutlierFinder::get_threshold() {
|
|
|
|
return _threshold;
|
|
|
|
}
|
|
|
|
void MLPPOutlierFinder::set_threshold(real_t val) {
|
|
|
|
_threshold = val;
|
|
|
|
}
|
|
|
|
|
|
|
|
Vector<Vector<real_t>> MLPPOutlierFinder::model_set_test(const Ref<MLPPMatrix> &input_set) {
|
|
|
|
ERR_FAIL_COND_V(!input_set.is_valid(), Vector<Vector<real_t>>());
|
|
|
|
|
|
|
|
MLPPStat stat;
|
|
|
|
|
|
|
|
Size2i input_set_size = input_set->size();
|
2023-01-24 19:20:18 +01:00
|
|
|
|
2023-02-09 15:30:33 +01:00
|
|
|
Vector<Vector<real_t>> outliers;
|
|
|
|
outliers.resize(input_set_size.y);
|
|
|
|
|
|
|
|
Ref<MLPPVector> input_set_i_row_tmp;
|
|
|
|
input_set_i_row_tmp.instance();
|
|
|
|
input_set_i_row_tmp->resize(input_set_size.x);
|
|
|
|
|
|
|
|
for (int i = 0; i < input_set_size.y; ++i) {
|
2023-04-29 15:07:30 +02:00
|
|
|
input_set->row_get_into_mlpp_vector(i, input_set_i_row_tmp);
|
2023-02-09 15:30:33 +01:00
|
|
|
real_t meanv = stat.meanv(input_set_i_row_tmp);
|
|
|
|
real_t s_dev_v = stat.standard_deviationv(input_set_i_row_tmp);
|
|
|
|
|
|
|
|
for (int j = 0; j < input_set_size.x; ++j) {
|
2023-04-29 13:44:18 +02:00
|
|
|
real_t input_set_i_j = input_set->element_get(i, j);
|
2023-02-09 15:30:33 +01:00
|
|
|
|
|
|
|
real_t z = (input_set_i_j - meanv) / s_dev_v;
|
|
|
|
|
|
|
|
if (ABS(z) > _threshold) {
|
|
|
|
outliers.write[i].push_back(input_set_i_j);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
return outliers;
|
2023-01-24 19:00:54 +01:00
|
|
|
}
|
2023-01-23 21:13:26 +01:00
|
|
|
|
2023-02-09 15:30:33 +01:00
|
|
|
Array MLPPOutlierFinder::model_set_test_bind(const Ref<MLPPMatrix> &input_set) {
|
|
|
|
Vector<Vector<real_t>> res = model_set_test(input_set);
|
|
|
|
|
|
|
|
Array arr;
|
|
|
|
|
|
|
|
for (int i = 0; i < res.size(); ++i) {
|
|
|
|
//will get converted to PoolRealArray
|
|
|
|
arr.push_back(Variant(res[i]));
|
|
|
|
}
|
|
|
|
|
|
|
|
return arr;
|
|
|
|
}
|
|
|
|
|
|
|
|
PoolVector2iArray MLPPOutlierFinder::model_set_test_indices(const Ref<MLPPMatrix> &input_set) {
|
|
|
|
ERR_FAIL_COND_V(!input_set.is_valid(), PoolVector2iArray());
|
|
|
|
|
|
|
|
MLPPStat stat;
|
|
|
|
|
|
|
|
Size2i input_set_size = input_set->size();
|
|
|
|
|
|
|
|
PoolVector2iArray outliers;
|
|
|
|
|
|
|
|
Ref<MLPPVector> input_set_i_row_tmp;
|
|
|
|
input_set_i_row_tmp.instance();
|
|
|
|
input_set_i_row_tmp->resize(input_set_size.x);
|
|
|
|
|
|
|
|
for (int i = 0; i < input_set_size.y; ++i) {
|
2023-04-29 15:07:30 +02:00
|
|
|
input_set->row_get_into_mlpp_vector(i, input_set_i_row_tmp);
|
2023-02-09 15:30:33 +01:00
|
|
|
real_t meanv = stat.meanv(input_set_i_row_tmp);
|
|
|
|
real_t s_dev_v = stat.standard_deviationv(input_set_i_row_tmp);
|
|
|
|
|
|
|
|
for (int j = 0; j < input_set_size.x; ++j) {
|
2023-04-29 13:44:18 +02:00
|
|
|
real_t z = (input_set->element_get(i, j) - meanv) / s_dev_v;
|
2023-02-09 15:30:33 +01:00
|
|
|
|
|
|
|
if (ABS(z) > _threshold) {
|
|
|
|
outliers.push_back(Vector2i(j, i));
|
2023-01-24 19:00:54 +01:00
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
2023-02-09 15:30:33 +01:00
|
|
|
|
2023-01-24 19:00:54 +01:00
|
|
|
return outliers;
|
|
|
|
}
|
2023-01-23 21:13:26 +01:00
|
|
|
|
2023-02-09 15:30:33 +01:00
|
|
|
PoolRealArray MLPPOutlierFinder::model_test(const Ref<MLPPVector> &input_set) {
|
|
|
|
ERR_FAIL_COND_V(!input_set.is_valid(), PoolRealArray());
|
|
|
|
|
|
|
|
MLPPStat stat;
|
|
|
|
PoolRealArray outliers;
|
|
|
|
|
|
|
|
real_t mean = stat.meanv(input_set);
|
|
|
|
real_t s_dev = stat.standard_deviationv(input_set);
|
|
|
|
|
|
|
|
int input_set_size = input_set->size();
|
|
|
|
const real_t *input_set_ptr = input_set->ptr();
|
|
|
|
|
|
|
|
for (int i = 0; i < input_set_size; ++i) {
|
|
|
|
real_t input_set_i = input_set_ptr[i];
|
|
|
|
|
|
|
|
real_t z = (input_set_i - mean) / s_dev;
|
|
|
|
|
|
|
|
if (ABS(z) > _threshold) {
|
|
|
|
outliers.push_back(input_set_i);
|
2023-01-24 19:00:54 +01:00
|
|
|
}
|
|
|
|
}
|
2023-02-09 15:30:33 +01:00
|
|
|
|
2023-01-24 19:00:54 +01:00
|
|
|
return outliers;
|
|
|
|
}
|
2023-02-09 15:30:33 +01:00
|
|
|
|
|
|
|
MLPPOutlierFinder::MLPPOutlierFinder(real_t threshold) {
|
|
|
|
_threshold = threshold;
|
|
|
|
}
|
|
|
|
|
|
|
|
MLPPOutlierFinder::MLPPOutlierFinder() {
|
|
|
|
_threshold = 0;
|
|
|
|
}
|
|
|
|
MLPPOutlierFinder::~MLPPOutlierFinder() {
|
|
|
|
}
|
|
|
|
|
|
|
|
void MLPPOutlierFinder::_bind_methods() {
|
|
|
|
ClassDB::bind_method(D_METHOD("get_threshold"), &MLPPOutlierFinder::get_threshold);
|
|
|
|
ClassDB::bind_method(D_METHOD("set_threshold", "val"), &MLPPOutlierFinder::set_threshold);
|
|
|
|
ADD_PROPERTY(PropertyInfo(Variant::REAL, "threshold"), "set_threshold", "get_threshold");
|
|
|
|
|
|
|
|
ClassDB::bind_method(D_METHOD("model_set_test", "input_set"), &MLPPOutlierFinder::model_set_test_bind);
|
|
|
|
ClassDB::bind_method(D_METHOD("model_set_test_indices", "input_set"), &MLPPOutlierFinder::model_set_test_indices);
|
|
|
|
|
|
|
|
ClassDB::bind_method(D_METHOD("model_test", "input_set"), &MLPPOutlierFinder::model_test);
|
|
|
|
}
|