expm1

auto kyosu::expm1 = eve::functor<expm1_t>

inlineconstexpr

Computes the exponential of the argument minus 1.

Header file

#include <kyosu/functions.hpp>

Callable Signatures

namespace kyosu
{
   template<kyosu::concepts::cayley_dickson T> constexpr complexify<T> expm1(T z)            noexcept; //1
   template<kyosu::concepts::cayley_dickson T> constexpr complexify<T> expm1[radpi](T z)     noexcept; //2
   template<kyosu::concepts::real           T> constexpr T             expm1[real_only](T z) noexcept; //3
}

Parameters

• z: Value to process.

Return value

1. Returns the exponential of the argument minus 1. Provisions are made to ensure good precision near zero.
2. z is multiplied by \(pi\) on input
3. Equivalent to eve::expm1 call.

Note

As it is a function devoted to better precision, it admits no raw option

Example

#include <eve/wide.hpp>
#include <iostream>
#include <kyosu/kyosu.hpp>
 
int main()
{
  using kyosu::complex_t;
  using kyosu::expm1;
  using kyosu::quaternion_t;
  using e_t = float;
  using c_t = kyosu::complex_t<float>;
  using q_t = kyosu::quaternion_t<float>;
  using we_t = eve::wide<e_t, eve::fixed<2>>;
  using wc_t = eve::wide<c_t, eve::fixed<2>>;
  using wq_t = eve::wide<q_t, eve::fixed<2>>;
 
  std::cout << "Real:        \n";
  e_t e(2.9f);
  we_t we = we_t(e);
  std::cout << e << " -> " << expm1(e) << "\n";
  std::cout << we << " -> " << expm1(we) << "\n";
  std::cout << expm1(c_t(e)) << "\n";
  std::cout << expm1(q_t(e)) << "\n";
  std::cout << expm1(wc_t(e)) << "\n";
  std::cout << expm1(wq_t(e)) << "\n";
 
  std::cout << "Complex:     \n";
  c_t c(3.5f, -2.9f);
  wc_t wc = wc_t(c);
  std::cout << c << " -> " << expm1(c) << "\n";
  std::cout << wc << " -> " << expm1(wc) << "\n";
  std::cout << expm1(q_t(c)) << "\n";
  std::cout << expm1(wq_t(c)) << "\n";
 
  std::cout << "Quaternion:  \n";
  q_t q(3.5f, -2.9f, 2.1f, 3.2f);
  wq_t wq = wq_t(q);
  std::cout << q << " -> " << expm1(q) << "\n";
  std::cout << wq << " -> " << expm1(wq) << "\n";
 
  return 0;
}