dc_scaledsec::modulo Interface Reference

Public Member Functions
type(dc_scaled_sec) function	dcscaledsec_modulo_si (sclsec, factor)
type(dc_scaled_sec) function	dcscaledsec_modulo_sr (sclsec, factor)
type(dc_scaled_sec) function	dcscaledsec_modulo_sd (sclsec, factor)
type(dc_scaled_sec) function	dcscaledsec_modulo_ss (sclsec, factor)

Detailed Description

Definition at line 228 of file dc_scaledsec.f90.

Member Function/Subroutine Documentation

dcscaledsec_modulo_sd()

type(dc_scaled_sec) function dc_scaledsec::modulo::dcscaledsec_modulo_sd	(	type(dc_scaled_sec), intent(in)	sclsec,
		real(dp), intent(in)	factor
	)

DC_SCALED_SEC を倍精度実数で割った剰余を計算

Parameters

[in]	sclsec	DC_SCALED_SEC 型変数 (被除数)
[in]	factor	倍精度実数の除数

Returns

除算の剰余

Definition at line 2402 of file dc_scaledsec.f90.

    use dc_message, only: messagenotify
    implicit none
    type(DC_SCALED_SEC), intent(in):: sclsec
    real(DP), intent(in):: factor
    type(DC_SCALED_SEC):: factor_scl
 
  continue
    factor_scl = factor
    result = modulo( sclsec, factor_scl )

dcscaledsec_modulo_si()

type(dc_scaled_sec) function dc_scaledsec::modulo::dcscaledsec_modulo_si	(	type(dc_scaled_sec), intent(in)	sclsec,
		integer, intent(in)	factor
	)

DC_SCALED_SEC を整数で割った剰余を計算

Parameters

[in]	sclsec	DC_SCALED_SEC 型変数 (被除数)
[in]	factor	整数の除数

Returns

除算の剰余

Definition at line 2348 of file dc_scaledsec.f90.

    use dc_message, only: messagenotify
    implicit none
    type(DC_SCALED_SEC), intent(in):: sclsec
    integer, intent(in):: factor
    type(DC_SCALED_SEC):: factor_scl
 
  continue
    factor_scl = factor
    result = modulo( sclsec, factor_scl )

dcscaledsec_modulo_sr()

type(dc_scaled_sec) function dc_scaledsec::modulo::dcscaledsec_modulo_sr	(	type(dc_scaled_sec), intent(in)	sclsec,
		real, intent(in)	factor
	)

DC_SCALED_SEC を単精度実数で割った剰余を計算

Parameters

[in]	sclsec	DC_SCALED_SEC 型変数 (被除数)
[in]	factor	単精度実数の除数

Returns

除算の剰余

Definition at line 2375 of file dc_scaledsec.f90.

    use dc_message, only: messagenotify
    implicit none
    type(DC_SCALED_SEC), intent(in):: sclsec
    real, intent(in):: factor
    type(DC_SCALED_SEC):: factor_scl
 
  continue
    factor_scl = factor
    result = modulo( sclsec, factor_scl )

dcscaledsec_modulo_ss()

type(dc_scaled_sec) function dc_scaledsec::modulo::dcscaledsec_modulo_ss	(	type(dc_scaled_sec), intent(in)	sclsec,
		type(dc_scaled_sec), intent(in)	factor
	)

DC_SCALED_SEC を割った際の剰余を計算

注意: Fortran コンパイラの精度制限により, factor は 10^12 未満である必要があります.

mod との違い: modulo は除数が正の場合、常に非負の結果を返します (Fortran の modulo セマンティクスに従います).

Parameters

[in]	sclsec	DC_SCALED_SEC 型変数 (被除数)
[in]	factor	DC_SCALED_SEC 型変数 (除数)

Returns

除算の剰余

Definition at line 2229 of file dc_scaledsec.f90.

    use dc_message, only: messagenotify
    use dc_types, only: dp_eps
    implicit none
    type(DC_SCALED_SEC), intent(in):: sclsec, factor
 
    type(DC_SCALED_SEC):: factor_scl
    real(DP):: sec_ary_mod(imin+imin:imax)
    integer:: i, move_down_index, sf_idx
    real(DP):: move_down
    real(DP):: factor_dp
    type(DC_SCALED_SEC):: zero_sec
    logical:: done
  continue
 
    ! factor must be smaller than 10^12 due to compiler precision limitations.
    !
    if ( .not. all( factor % sec_ary (imax-4:imax) == (/ 0, 0, 0, 0, 0 /) ) ) then
      call messagenotify( 'E', 'dc_scaledsec#modulo', &
        & 'factor must be smaller than 10^12' )
    end if
 
    if ( sclsec == factor ) then
      result = zero_sec
      return
    end if
 
    factor_scl % sec_ary(imin:-1) = 0
    factor_scl % sec_ary(imin-imin:imax) = factor % sec_ary(imin:imax+imin)
    factor_scl % flag_negative = factor % flag_negative
 
    factor_dp = factor_scl
 
    done = .false.
    move_down = 0.0_dp
    do i = imax, imin, -1
      move_down_index = i
      if ( abs(move_down) > dp_eps ) then
        sf_idx = i - imin
        if ( sf_idx > ubound(scale_factor_xx, 1) ) sf_idx = ubound(scale_factor_xx, 1)
        if ( sf_idx < lbound(scale_factor_xx, 1) ) sf_idx = lbound(scale_factor_xx, 1)
        if ( abs(factor_dp) > ( move_down + scale_factor ) * scale_factor_xx( sf_idx ) ) then
          done = .true.
          exit
        end if
      end if
 
      sec_ary_mod(i) = &
        & mod( ( sclsec % sec_ary(i) + move_down ), factor_dp )
 
      if ( abs(sec_ary_mod(i)) > dp_eps ) then
        move_down = sec_ary_mod(i) * scale_factor
      else
        move_down = 0.0_dp
      end if
    end do
 
    if ( .not. done ) then
      do i = imin - 1, imin + imin, -1
      move_down_index = i
      if ( abs(move_down) > dp_eps ) then
        sf_idx = i - imin
        if ( sf_idx > ubound(scale_factor_xx, 1) ) sf_idx = ubound(scale_factor_xx, 1)
        if ( sf_idx < lbound(scale_factor_xx, 1) ) sf_idx = lbound(scale_factor_xx, 1)
        if ( abs(factor_dp) > ( move_down + scale_factor ) * scale_factor_xx( sf_idx ) ) then
          done = .true.
          exit
        end if
      end if
 
      sec_ary_mod(i) = mod( move_down, factor_dp )
 
      if ( abs(sec_ary_mod(i)) > dp_eps ) then
        move_down = sec_ary_mod(i) * scale_factor
      else
        move_down = 0.0_dp
      end if
      end do
    end if
 
    result = move_down * scale_factor_xx(move_down_index)
    if ( move_down_index > imin - 1 ) then
      result % sec_ary(imin:move_down_index) = sclsec % sec_ary(imin:move_down_index)
    end if
    
    result % flag_negative = .false.
 
    if ( .not. result == zero_sec ) then
      if ( .not. sclsec % flag_negative .and. factor % flag_negative ) then
        result = - factor - result
        result % flag_negative = .not. sclsec % flag_negative
        
      elseif ( sclsec % flag_negative .and. .not. factor % flag_negative ) then
        result = factor - result
        result % flag_negative = .not. sclsec % flag_negative
 
      else
        result % flag_negative = sclsec % flag_negative
        
      end if
    end if
 

References dc_types::dp_eps.

---

The documentation for this interface was generated from the following file:

• dc_utils/dc_scaledsec.f90