ROperator_BasicUnary.hxx

Go to the documentation of this file.

#ifndef TMVA_EXPERIMENTAL_SOFIE_ROPERATOR_BASIC_UNARY
#define TMVA_EXPERIMENTAL_SOFIE_ROPERATOR_BASIC_UNARY
 
#include <TMVA/ROperator.hxx>
#include <TMVA/RModel.hxx>
#include <TMVA/SOFIE_common.hxx>
 
namespace TMVA {
namespace Experimental {
namespace SOFIE {
 
enum class EBasicUnaryOperator { kReciprocal, kSqrt , kNeg, kExp, kLog };
 
template <typename T, EBasicUnaryOperator Op>
struct UnaryOpTraits {
};
 
template <typename T>
struct UnaryOpTraits<T, EBasicUnaryOperator::kReciprocal> {
   static std::string Name() { return "Reciprocal"; }
   static std::string Op(const std::string &X) { return "1/" + X; }
};
 
template <typename T>
struct UnaryOpTraits<T, EBasicUnaryOperator::kSqrt> {
   static std::string Name() { return "Sqrt"; }
   static std::string Op(const std::string &X) { return "std::sqrt(" + X + ")"; }
};
 
template <typename T>
struct UnaryOpTraits<T, EBasicUnaryOperator::kNeg> {
   static std::string Name() { return "Neg"; }
   static std::string Op(const std::string &X) { return "-" + X; }
};
 
template <typename T>
struct UnaryOpTraits<T, EBasicUnaryOperator::kExp> {
   static std::string Name() { return "Exp"; }
   static std::string Op(const std::string &X) { return "std::exp(" + X + ")"; }
};
 
template <typename T>
struct UnaryOpTraits<T, EBasicUnaryOperator::kLog> {
   static std::string Name() { return "Log"; }
   static std::string Op(const std::string &X) { return "std::log(" + X + ")"; }
};
 
template <typename T, EBasicUnaryOperator Op>
class ROperator_BasicUnary final : public ROperator {
private:
   std::string fNX;
   std::string fNY;
 
   std::vector<size_t> fShapeX;
   std::vector<size_t> fShapeY;
 
public:
   ROperator_BasicUnary() {}
 
   ROperator_BasicUnary(std::string nameX, std::string nameY)
      : fNX(UTILITY::Clean_name(nameX)), fNY(UTILITY::Clean_name(nameY))
   {}
 
   std::vector<std::vector<size_t>> ShapeInference(std::vector<std::vector<size_t>> input) override { return input; }
 
   std::vector<ETensorType> TypeInference(std::vector<ETensorType> input) override { return input; }
 
   void Initialize(RModel &model) override
   {
      if (!model.CheckIfTensorAlreadyExist(fNX)) {
         throw std::runtime_error("TMVA::SOFIE - Tensor " + fNX + " not found.");
      }
      fShapeX = model.GetTensorShape(fNX);
      fShapeY = ShapeInference({fShapeX})[0];
      model.AddIntermediateTensor(fNY, model.GetTensorType(fNX), fShapeY);
   }
 
   std::string Generate(std::string OpName) override
   {
      OpName = "op_" + OpName;
      std::stringstream out;
 
      out << SP << "\n//---- Operator" << UnaryOpTraits<T, Op>::Name() << " " << OpName << "\n";
      size_t length = ConvertShapeToLength(fShapeX);
      out << SP << "for (size_t i = 0; i < " << length << "; i++) {\n";
      out << SP << SP << "tensor_" << fNY << "[i] = " << UnaryOpTraits<T, Op>::Op("tensor_" + fNX + "[i]") << ";\n";
      out << SP << "}\n";
      return out.str();
   }
 
   std::vector<std::string> GetStdLibs() override {
      if (Op == EBasicUnaryOperator::kSqrt || Op == EBasicUnaryOperator::kExp || Op == EBasicUnaryOperator::kLog) {
         return { std::string("cmath") };
      } else {
         return {};
      }
   }
};
 
} // namespace SOFIE
} // namespace Experimental
} // namespace TMVA
 
#endif