Changes to pamela

This file lists changes made to pamela, including bug fixes and upgrades, dating from August 2000. T.R.Marsh.

Date File Changes made

04/11/08 Version 1.0.3 A few fixes made to some scripts, including the 'fixhead' for WHT 1998 fast CCD data.

10/04/08 Version 1.0.2 Version release to allow for changes made to many scripts and minor tweaks to Makefile. I expect that there are still some scripts that need updating.

22/04/08 Many scripts Having to put changes in for new arbitrary location of starlink code.

10/04/08 Version 1.0.1 Lots of changes for the latest starlink, and I have fully moved to g95 and now link to the native version of pgplot as provided by starlink.

18/10/07 Version 1.0.0 At long last I have added a version number to pamela.

16/10/07 fixhead.pl Added new formats from John for Magellan data.

03/10/07 track.f Bug spotted by Pierre. Sort of bug that would either prevent track working altogether or do nothing.

17/07/07 picstat.f Fixed irritating shift between gaussian and histogram

17/07/07 Makefile Changed Pamela to pamela_aliases to avoid a problem with MAC's case-insensitive file system.

17/02/06 reduce.pl Added options to allow users to edit the file of data types if idtype gets them wrong. -t allows you to run reduce, save the dattypes and then halt the reduction, while -s allows you to skip the data typing and instead read in whatever file was saved (and possibly edited) from the -t run.

12/12/05 lsnosdf.csh
reduce.pl New script to list sdf files without their trailing .sdfs. Update to reduce to allow fixed bad pixels to be eliminated from the sky regions.

10/12/05 Many scripts I have made many improvments to various scripts. The main one has been to remove all format dependent stuff from 'reduce.pl' and put it in a single place, 'fixhead.pl'. The other is to add an option in reduce to support measurement of the readout noise on each frame as I found that the Marconi chip on ISIS had a variable level of read noise.

24/01/05 track Improvment to fitting in track to make it less sensitive to high weight but poor points.

21/01/05 reduce
profit
extopt I discovered something that may be well know to others which is that profit has a hard time for faint objects under load of cosmic rays. The reason is that under these circumstances, rows with cosmic rays tend to get overweighted. In an attempt to improve this, I have added the option of median filtering the raw sums across the profiles that are used when estimating the fractions. The idea is that the cosmic rays will get kicked out and so the raw estimates of the fractions will be nearer the mark. The cosmic rays themselves will then stand out more and will be kicked out during the fits. It definitely seems to be an improvement.

14/01/05 uniqformat
makebias
listformat
debias
loopdisplay
whtdiv Stack of new scripts and changed scripts.

13/04/03 profit.f Added a one-by-one bad pixel rejection option. This is slow but needed in rare pathological cases. Computers are so much faster now that it is at least a possibility.

13/04/03 fixhead Will not now immediately bomb out if it can't find something, but goes onto next frame. Puts default values into any frames it can't manage.

13/04/03 splfit Fixed this which appeared to be terminally broken.

21/03/02 Makefile Now defines command search path so that you do not have to have run the STARLINK scripts first.

21/03/02 genhtml.pl Now writes an INDEX.html file automatically.

04/12/01 makefile Modified to improve setup procedure, mainly to allow multiple setups.

30/12/01 makefile Modified to make setup more self-contained without links back to a directory that could disappear. Script called "debias" is automatically modified to reflect the installation directory.

02/08/01 track.f Fairly major improvement to tweak option to stop it being too picky. What was happening was that points were being rejected on the tweak because they had moved more than the maximum allowed, but this is not what is wanted on a tweak except on later cycles. Much more robust now. Requires an extra parameter to constrain the shift on the first measuring cycle separately from later ones.

02/08/01 reduce
datafile More mods to these files.

02/08/01 loganal Corrected and upgraded this analysis script. The correction was to make it independent of the specific installation of pamela.

01/08/01 reduce
datafile Extensive changes/improvements made to the reduce script.

27/07/01 makefile A change of the pamela installation. Now sticks executable, object library and compiled code in a single directory. This is to allow for multiple versions since I live under two operating systems.

06/04/01 extopt.f
extnor.f
optext.f Another attempt at removing error messages when using non-standard sky files (e.g. zero files)

13/03/01 noise.f Changed plot labels

06/02/01 makefile Changed to a more generic form to accomodate non-gmake make.

06/02/01 makefile Spotted an argument mistype bug in track.f

06/02/01 makefile Too many arguments to PGOPEN in extopt.f

01/01/01 fixhead.pl This script was not correcting the UTs to the centre of the exposures!

11/08/00 extnor.f Fixed a bug that was preventing prompting of ystart and yend, due to darkframe.

10/08/00 loganal.pl New script to examine contents of reduction log files which can be very large. This script strips out main indicators of problems.

04/08/00 profit.f Extended the region over which bad pixels are tested for when checking which blocks are OK. This is because later on a more extended X range is used which can pick up bad pixels not noticed before.

03/08/00 makefile Now updates location of pamela in the debias.pl script. Have also added the Perl directory to the tar file to get the documentation perl script.

03/08/00 scripts Have updated their documentation for new document.pl script. Also added ./ to the calls to system inside them to avoid users having to rely on the path.

03/08/00 polfit.f Fixed some bugs caused by the possibility of bad data spotted by Pierre.

Date	File	Changes made
04/11/08	Version 1.0.3	A few fixes made to some scripts, including the 'fixhead' for WHT 1998 fast CCD data.
10/04/08	Version 1.0.2	Version release to allow for changes made to many scripts and minor tweaks to Makefile. I expect that there are still some scripts that need updating.
22/04/08	Many scripts	Having to put changes in for new arbitrary location of starlink code.
10/04/08	Version 1.0.1	Lots of changes for the latest starlink, and I have fully moved to g95 and now link to the native version of pgplot as provided by starlink.
18/10/07	Version 1.0.0	At long last I have added a version number to pamela.
16/10/07	fixhead.pl	Added new formats from John for Magellan data.
03/10/07	track.f	Bug spotted by Pierre. Sort of bug that would either prevent track working altogether or do nothing.
17/07/07	picstat.f	Fixed irritating shift between gaussian and histogram
17/07/07	Makefile	Changed Pamela to pamela_aliases to avoid a problem with MAC's case-insensitive file system.
17/02/06	reduce.pl	Added options to allow users to edit the file of data types if idtype gets them wrong. -t allows you to run reduce, save the dattypes and then halt the reduction, while -s allows you to skip the data typing and instead read in whatever file was saved (and possibly edited) from the -t run.
12/12/05	lsnosdf.csh reduce.pl	New script to list sdf files without their trailing .sdfs. Update to reduce to allow fixed bad pixels to be eliminated from the sky regions.
10/12/05	Many scripts	I have made many improvments to various scripts. The main one has been to remove all format dependent stuff from 'reduce.pl' and put it in a single place, 'fixhead.pl'. The other is to add an option in reduce to support measurement of the readout noise on each frame as I found that the Marconi chip on ISIS had a variable level of read noise.
24/01/05	track	Improvment to fitting in track to make it less sensitive to high weight but poor points.
21/01/05	reduce profit extopt	I discovered something that may be well know to others which is that profit has a hard time for faint objects under load of cosmic rays. The reason is that under these circumstances, rows with cosmic rays tend to get overweighted. In an attempt to improve this, I have added the option of median filtering the raw sums across the profiles that are used when estimating the fractions. The idea is that the cosmic rays will get kicked out and so the raw estimates of the fractions will be nearer the mark. The cosmic rays themselves will then stand out more and will be kicked out during the fits. It definitely seems to be an improvement.
14/01/05	uniqformat makebias listformat debias loopdisplay whtdiv	Stack of new scripts and changed scripts.
13/04/03	profit.f	Added a one-by-one bad pixel rejection option. This is slow but needed in rare pathological cases. Computers are so much faster now that it is at least a possibility.
13/04/03	fixhead	Will not now immediately bomb out if it can't find something, but goes onto next frame. Puts default values into any frames it can't manage.
13/04/03	splfit	Fixed this which appeared to be terminally broken.
21/03/02	Makefile	Now defines command search path so that you do not have to have run the STARLINK scripts first.
21/03/02	genhtml.pl	Now writes an INDEX.html file automatically.
04/12/01	makefile	Modified to improve setup procedure, mainly to allow multiple setups.
30/12/01	makefile	Modified to make setup more self-contained without links back to a directory that could disappear. Script called "debias" is automatically modified to reflect the installation directory.
02/08/01	track.f	Fairly major improvement to tweak option to stop it being too picky. What was happening was that points were being rejected on the tweak because they had moved more than the maximum allowed, but this is not what is wanted on a tweak except on later cycles. Much more robust now. Requires an extra parameter to constrain the shift on the first measuring cycle separately from later ones.
02/08/01	reduce datafile	More mods to these files.
02/08/01	loganal	Corrected and upgraded this analysis script. The correction was to make it independent of the specific installation of pamela.
01/08/01	reduce datafile	Extensive changes/improvements made to the reduce script.
27/07/01	makefile	A change of the pamela installation. Now sticks executable, object library and compiled code in a single directory. This is to allow for multiple versions since I live under two operating systems.
06/04/01	extopt.f extnor.f optext.f	Another attempt at removing error messages when using non-standard sky files (e.g. zero files)
13/03/01	noise.f	Changed plot labels
06/02/01	makefile	Changed to a more generic form to accomodate non-gmake make.
06/02/01	makefile	Spotted an argument mistype bug in track.f
06/02/01	makefile	Too many arguments to PGOPEN in extopt.f
01/01/01	fixhead.pl	This script was not correcting the UTs to the centre of the exposures!
11/08/00	extnor.f	Fixed a bug that was preventing prompting of ystart and yend, due to darkframe.
10/08/00	loganal.pl	New script to examine contents of reduction log files which can be very large. This script strips out main indicators of problems.
04/08/00	profit.f	Extended the region over which bad pixels are tested for when checking which blocks are OK. This is because later on a more extended X range is used which can pick up bad pixels not noticed before.
03/08/00	makefile	Now updates location of pamela in the debias.pl script. Have also added the Perl directory to the tar file to get the documentation perl script.
03/08/00	scripts	Have updated their documentation for new document.pl script. Also added ./ to the calls to system inside them to avoid users having to rely on the path.
03/08/00	polfit.f	Fixed some bugs caused by the possibility of bad data spotted by Pierre.

Tom Marsh, Warwick