On Mon, 23 Jan 2006, Brad Cavanagh wrote:
> On Mon, 23 Jan 2006, Tim Jenness wrote:
> Oh hold on a second. AUTOPHOTOM fails in the same way on Linux as for OS X.
> However, I was defining the X/Y position in GRID coordinates. If I define
> them in PIXEL (which changes the position from 139,761 to 8,627) it works on
> both Linux and OS X.
>
> However, I still can't get it to work through GAIA on OS X. It doesn't matter
> what I change the coordinate frame to, aperture photometry still fails.
>
> Yes, if I set the pixel origin to [0,0] aperture photometry works fine in
> GAIA.
So....does AUTOPHOTOM want the coordinates as PIXEL or GRID?
> ==31868== Conditional jump or move depends on uninitialised value(s)
> ==31868== at 0x1E3BA987: cnfCptr (cnfMem.c:192)
> ==31868== by 0x1E3BAAFB: cnf_pval_ (cnfMem.c:482)
> ==31868== by 0x804BCA4: autophotom_ (autophotom.f:618)
> ==31868== by 0x804B55D: photom_mon_ (photom_mon.f:67)
> ==31868==
> ==31868== Conditional jump or move depends on uninitialised value(s)
> ==31868== at 0x1E3BA987: cnfCptr (cnfMem.c:192)
> ==31868== by 0x1E3BAAFB: cnf_pval_ (cnfMem.c:482)
> ==31868== by 0x804BCBB: autophotom_ (autophotom.f:618)
> ==31868== by 0x804B55D: photom_mon_ (photom_mon.f:67)
These two are caused by IPVAR and IPMASK not being initialised to 0
properly.
> ==31868==
> ==31868== Use of uninitialised value of size 4
> ==31868== at 0x805F342: pho1_gcalc_ (pho1_gcalc.f:151)
> ==31868== by 0x805EB2D: pho1_auto_ (pho1_auto.f:1147)
> ==31868== by 0x804BD23: autophotom_ (autophotom.f:618)
> ==31868== by 0x804B55D: photom_mon_ (photom_mon.f:67)
> ==31868==
> ==31868== Use of uninitialised value of size 4
> ==31868== at 0x805F362: pho1_gcalc_ (pho1_gcalc.f:155)
> ==31868== by 0x805EB2D: pho1_auto_ (pho1_auto.f:1147)
> ==31868== by 0x804BD23: autophotom_ (autophotom.f:618)
> ==31868== by 0x804B55D: photom_mon_ (photom_mon.f:67)
ie
VSTAR = DBLE( LVSTAR )
VSKY = DBLE( LVSKY )
and can be fixed trivially if VSTAR and VSKY are only set from LVSTAR and
LVSKY if Variance is being used.
Brad: Are you going to fix these? :-)
--
Tim Jenness
JAC software
http://www.jach.hawaii.edu/~timj
|