rmontanana/BayesNet
1
0
Fork 0
Code Issues 2 Pull Requests 1 Actions Packages Projects Releases 6 Wiki Activity

Restart proposal

Browse Source
This commit is contained in:
Ricardo Montañana Gómez
2025-08-21 19:20:03 +02:00
parent 1aa3b609e5
commit 74b391907a
1 changed files with 14 additions and 20 deletions
Download Patch File Download Diff File Expand all files Collapse all files

20
bayesnet/classifiers/Proposal.cc
View File

@@ -118,21 +118,17 @@ namespace bayesnet {
} }
return states; return states;
} }
map<std::string, std::vector<int>> Proposal::fit_local_discretization(const torch::Tensor& y, map<std::string, std::vector<int>> states) map<std::string, std::vector<int>> Proposal::fit_local_discretization(const torch::Tensor& y, map<std::string, std::vector<int>> states_)
{ {
// Discretize the continuous input data and build pDataset (Classifier::dataset) // Discretize the continuous input data and build pDataset (Classifier::dataset)
// We expect to have in states for numeric features an empty vector and for discretized features a vector of states
int m = Xf.size(1); int m = Xf.size(1);
int n = Xf.size(0); int n = Xf.size(0);
map<std::string, std::vector<int>> states;
pDataset = torch::zeros({ n + 1, m }, torch::kInt32); pDataset = torch::zeros({ n + 1, m }, torch::kInt32);
auto yv = std::vector<int>(y.data_ptr<int>(), y.data_ptr<int>() + y.size(0)); auto yv = std::vector<int>(y.data_ptr<int>(), y.data_ptr<int>() + y.size(0));
// discretize input data by feature(row) // discretize input data by feature(row)
std::unique_ptr<mdlp::Discretizer> discretizer; std::unique_ptr<mdlp::Discretizer> discretizer;
for (auto i = 0; i < pFeatures.size(); ++i) { for (auto i = 0; i < pFeatures.size(); ++i) {
auto Xt_ptr = Xf.index({ i }).data_ptr<float>();
auto Xt = std::vector<float>(Xt_ptr, Xt_ptr + Xf.size(1));
if (states[pFeatures[i]].empty()) {
// If the feature is numeric, we discretize it
if (discretizationType == discretization_t::BINQ) { if (discretizationType == discretization_t::BINQ) {
discretizer = std::make_unique<mdlp::BinDisc>(ld_params.proposed_cuts, mdlp::strategy_t::QUANTILE); discretizer = std::make_unique<mdlp::BinDisc>(ld_params.proposed_cuts, mdlp::strategy_t::QUANTILE);
} else if (discretizationType == discretization_t::BINU) { } else if (discretizationType == discretization_t::BINU) {
@@ -140,15 +136,13 @@ namespace bayesnet {
} else { // Default is MDLP } else { // Default is MDLP
discretizer = std::make_unique<mdlp::CPPFImdlp>(ld_params.min_length, ld_params.max_depth, ld_params.proposed_cuts); discretizer = std::make_unique<mdlp::CPPFImdlp>(ld_params.min_length, ld_params.max_depth, ld_params.proposed_cuts);
} }
pDataset.index_put_({ i, "..." }, torch::tensor(discretizer->fit_transform(Xt, yv))); auto Xt_ptr = Xf.index({ i }).data_ptr<float>();
int n_states = discretizer->getCutPoints().size() + 1; auto Xt = std::vector<float>(Xt_ptr, Xt_ptr + Xf.size(1));
auto xStates = std::vector<int>(n_states); discretizer->fit(Xt, yv);
pDataset.index_put_({ i, "..." }, torch::tensor(discretizer->transform(Xt)));
auto xStates = std::vector<int>(discretizer->getCutPoints().size() + 1);
iota(xStates.begin(), xStates.end(), 0); iota(xStates.begin(), xStates.end(), 0);
states[pFeatures[i]] = xStates; states[pFeatures[i]] = xStates;
} else {
// If the feature is categorical, we just copy it
pDataset.index_put_({ i, "..." }, Xf[i].to(torch::kInt32));
}
discretizers[pFeatures[i]] = std::move(discretizer); discretizers[pFeatures[i]] = std::move(discretizer);
} }
int n_classes = torch::max(y).item<int>() + 1; int n_classes = torch::max(y).item<int>() + 1;