CLIP odd code

[DeadParrot]

Another odd looking bit of code from CLIP:

NVS = NABOVE + 2;
NumVertInShadowOrClippedSurface = NABOVE + 2;
XVT( NVT + 1 ) = XVT( 1 );
YVT( NVT + 1 ) = YVT( 1 );
ZVT( NVT + 1 ) = ZVT( 1 );
for ( N = 1; N <= NVT; ++N ) {
   if ( ZVT( N ) >= 0.0 && ZVT( N + 1 ) < 0.0 ) { // Line enters plane of receiving surface
      XIN = ( ZVT( N + 1 ) * XVT( N ) - ZVT( N ) * XVT( N + 1 ) ) / ( ZVT( N + 1 ) - ZVT( N ) );
      YIN = ( ZVT( N + 1 ) * YVT( N ) - ZVT( N ) * YVT( N + 1 ) ) / ( ZVT( N + 1 ) - ZVT( N ) );
   } else if ( ZVT( N ) <= 0.0 && ZVT( N + 1 ) > 0.0 ) { // Line exits plane of receiving surface
      NEXT = N + 1;
      XOUT = ( ZVT( N + 1 ) * XVT( N ) - ZVT( N ) * XVT( N + 1 ) ) / ( ZVT( N + 1 ) - ZVT( N ) );
      YOUT = ( ZVT( N + 1 ) * YVT( N ) - ZVT( N ) * YVT( N + 1 ) ) / ( ZVT( N + 1 ) - ZVT( N ) );
   }
}

It just keeps going on to next N after setting the IN or OUT coords, only keeping the last ones. It also uses the IN and OUT values even if they are never set. If the usage assures that the IN and OUT cases are satisfied exactly once over all N then I guess it is OK but that should really be documented and assert-checked.

Comments?

[rraustad]

Again needing help from the author....

IF (NABOVE+NON == NVT) THEN ! Rename the unclipped shadow casting
surface.

 NVS=NVT
 NumVertInShadowOrClippedSurface = NVT
 DO N = 1, NVT
   XVC(N) = XVT(N)
   YVC(N) = YVT(N)
   ZVC(N) = ZVT(N)
 END DO

ELSEIF (NABOVE == 0) THEN ! Totally submerged shadow casting surface.

 NVS=0
 NumVertInShadowOrClippedSurface = 0

ELSE ! Remove (clip) that portion of the shadow casting surface.
which is below the receiving surface

 NVS        = NABOVE + 2
 NumVertInShadowOrClippedSurface = NABOVE + 2
 XVT(NVT+1) = XVT(1)
 YVT(NVT+1) = YVT(1)
 ZVT(NVT+1) = ZVT(1)
 DO N = 1, NVT
   IF (ZVT(N) >= 0.0d0 .AND. ZVT(N+1) < 0.0d0) THEN ! Line enters 

plane of receiving surface

     XIN=(ZVT(N+1)*XVT(N)-ZVT(N)*XVT(N+1))/(ZVT(N+1)-ZVT(N))
     YIN=(ZVT(N+1)*YVT(N)-ZVT(N)*YVT(N+1))/(ZVT(N+1)-ZVT(N))

   ELSEIF (ZVT(N) <= 0.0d0 .AND. ZVT(N+1) > 0.0d0) THEN ! Line exits 

plane of receiving surface

     NEXT=N+1
     XOUT=(ZVT(N+1)*XVT(N)-ZVT(N)*XVT(N+1))/(ZVT(N+1)-ZVT(N))
     YOUT=(ZVT(N+1)*YVT(N)-ZVT(N)*YVT(N+1))/(ZVT(N+1)-ZVT(N))

   END IF
 END DO

       ! Renumber the vertices of the clipped shadow casting 

surface. so they are still counter-clockwise sequential.

 XVC(1)   = XOUT
 YVC(1)   = YOUT
 ZVC(1)   = 0.0d0
 XVC(NVS) = XIN
 YVC(NVS) = YIN
 ZVC(NVS) = 0.0d0
 DO N = 1, NABOVE
   IF (NEXT > NVT) NEXT = 1
   XVC(N+1) = XVT(NEXT)
   YVC(N+1) = YVT(NEXT)
   ZVC(N+1) = ZVT(NEXT)
   NEXT     = NEXT+1
 END DO

END IF

On 6/23/2014 6:18 PM, Stuart Mentzer wrote:

Another odd looking bit of code from CLIP:

NVS = NABOVE + 2;
NumVertInShadowOrClippedSurface = NABOVE + 2;
XVT( NVT + 1 ) = XVT( 1 );
YVT( NVT + 1 ) = YVT( 1 );
ZVT( NVT + 1 ) = ZVT( 1 );
for ( N = 1; N <= NVT; ++N ) {
if ( ZVT( N ) >= 0.0 && ZVT( N + 1 ) < 0.0 ) { // Line enters plane of receiving surface
XIN = ( ZVT( N + 1 ) * XVT( N ) - ZVT( N ) * XVT( N + 1 ) ) / ( ZVT( N + 1 ) - ZVT( N ) );
YIN = ( ZVT( N + 1 ) * YVT( N ) - ZVT( N ) * YVT( N + 1 ) ) / ( ZVT( N + 1 ) - ZVT( N ) );
} else if ( ZVT( N ) <= 0.0 && ZVT( N + 1 ) > 0.0 ) { // Line exits plane of receiving surface
NEXT = N + 1;
XOUT = ( ZVT( N + 1 ) * XVT( N ) - ZVT( N ) * XVT( N + 1 ) ) / ( ZVT( N + 1 ) - ZVT( N ) );
YOUT = ( ZVT( N + 1 ) * YVT( N ) - ZVT( N ) * YVT( N + 1 ) ) / ( ZVT( N + 1 ) - ZVT( N ) );
}
}

It just keeps going on to next N after setting the IN or OUT coords,
only keeping the last ones. It also uses the IN and OUT values even if
they are never set. If the usage assures that the IN and OUT cases are
satisfied exactly once over all N then I guess it is OK but that
should really be documented and assert-checked.

Comments?

—
Reply to this email directly or view it on GitHub
#4315.

Richard Raustad
Senior Research Engineer
Florida Solar Energy Center
1679 Clearlake Road
Cocoa, FL 32922
Ph: (321)638-1454
http://www.fsec.ucf.edu/en/

Electric Vehicle Transportation Center
http://evtc.fsec.ucf.edu/

[jmarrec]

EnergyPlus/src/EnergyPlus/SolarShading.cc

Lines 3306 to 3325 in 0c9b8c4

	state.dataSolarShading->NVS = NABOVE + 2;
	state.dataSolarShading->NumVertInShadowOrClippedSurface = NABOVE + 2;
	Real64 ZVT_N, ZVT_P(ZVT(1));
	XVT(NVT + 1) = XVT(1);
	YVT(NVT + 1) = YVT(1);
	ZVT(NVT + 1) = ZVT_P;
	for (int N = 1, P = 2; N <= NVT; ++N, ++P) {
	ZVT_N = ZVT_P;
	ZVT_P = ZVT(P);
	if (ZVT_N >= 0.0 && ZVT_P < 0.0) { // Line enters plane of receiving surface
	Real64 const ZVT_fac(1.0 / (ZVT_P - ZVT_N));
	XIN = (ZVT_P * XVT(N) - ZVT_N * XVT(P)) * ZVT_fac;
	YIN = (ZVT_P * YVT(N) - ZVT_N * YVT(P)) * ZVT_fac;
	} else if (ZVT_N <= 0.0 && ZVT_P > 0.0) { // Line exits plane of receiving surface
	NEXT = N + 1;
	Real64 const ZVT_fac(1.0 / (ZVT_P - ZVT_N));
	XOUT = (ZVT_P * XVT(N) - ZVT_N * XVT(P)) * ZVT_fac;
	YOUT = (ZVT_P * YVT(N) - ZVT_N * YVT(P)) * ZVT_fac;
	}
	}