Line data    Source code

       1              : // ***************************************************************
       2              : // SPDX-FileCopyrightText: Copyright 2024 Ricardo Montañana Gómez
       3              : // SPDX-FileType: SOURCE
       4              : // SPDX-License-Identifier: MIT
       5              : // ***************************************************************
       6              : 
       7              : #include "SPODELd.h"
       8              : 
       9              : namespace bayesnet {
      10           55 :     SPODELd::SPODELd(int root) : SPODE(root), Proposal(dataset, features, className) {}
      11           42 :     SPODELd& SPODELd::fit(torch::Tensor& X_, torch::Tensor& y_, const std::vector<std::string>& features_, const std::string& className_, map<std::string, std::vector<int>>& states_)
      12              :     {
      13           42 :         checkInput(X_, y_);
      14           42 :         Xf = X_;
      15           42 :         y = y_;
      16           42 :         return commonFit(features_, className_, states_);
      17              :     }
      18              : 
      19            2 :     SPODELd& SPODELd::fit(torch::Tensor& dataset, const std::vector<std::string>& features_, const std::string& className_, map<std::string, std::vector<int>>& states_)
      20              :     {
      21            2 :         if (!torch::is_floating_point(dataset)) {
      22            1 :             throw std::runtime_error("Dataset must be a floating point tensor");
      23              :         }
      24            4 :         Xf = dataset.index({ torch::indexing::Slice(0, dataset.size(0) - 1), "..." }).clone();
      25            3 :         y = dataset.index({ -1, "..." }).clone().to(torch::kInt32);
      26            1 :         return commonFit(features_, className_, states_);
      27            3 :     }
      28              : 
      29           43 :     SPODELd& SPODELd::commonFit(const std::vector<std::string>& features_, const std::string& className_, map<std::string, std::vector<int>>& states_)
      30              :     {
      31           43 :         features = features_;
      32           43 :         className = className_;
      33              :         // Fills std::vectors Xv & yv with the data from tensors X_ (discretized) & y
      34           43 :         states = fit_local_discretization(y);
      35              :         // We have discretized the input data
      36              :         // 1st we need to fit the model to build the normal SPODE structure, SPODE::fit initializes the base Bayesian network
      37           43 :         SPODE::fit(dataset, features, className, states);
      38           43 :         states = localDiscretizationProposal(states, model);
      39           43 :         return *this;
      40              :     }
      41           34 :     torch::Tensor SPODELd::predict(torch::Tensor& X)
      42              :     {
      43           34 :         auto Xt = prepareX(X);
      44           68 :         return SPODE::predict(Xt);
      45           34 :     }
      46            9 :     std::vector<std::string> SPODELd::graph(const std::string& name) const
      47              :     {
      48            9 :         return SPODE::graph(name);
      49              :     }
      50              : }