Lenstool issueshttps://git-cral.univ-lyon1.fr/lenstool/lenstool/-/issues2023-08-06T09:30:20Zhttps://git-cral.univ-lyon1.fr/lenstool/lenstool/-/issues/6Compiling issue in release-8: missing LDFLAGS2023-08-06T09:30:20ZBenjamin BeauchesneCompiling issue in release-8: missing LDFLAGSThere seem to be some missing default flags with the addition of libtool. This problem happened to a collaborator who did not have general LDFLAGS set up. In my case, there are provided through my conda environment and are the following:...There seem to be some missing default flags with the addition of libtool. This problem happened to a collaborator who did not have general LDFLAGS set up. In my case, there are provided through my conda environment and are the following:
-Wl,-pie -Wl,-headerpad_max_install_names -Wl,-dead_strip_dylibs -Wl,-rpath,/path/to/miniconda3/lib -L/path/to/miniconda3/lib
With there are no issues, but without them, my collaborator ran into compiling problems on Linux. I will try to update the issue with the error that he got. If I removed them, I could compile the code fine on my MacBook with clang version 11.1.0 but if I run the binaries, there is the following error:
dyld: Library not loaded: @rpath/libgsl.25.dylib
Referenced from: /path/to/the/binary
Reason: image not found
Abort trap: 6Eric JulloEric Jullohttps://git-cral.univ-lyon1.fr/lenstool/lenstool/-/issues/4Issue in the makefile in release-82023-06-07T13:48:24ZGuillaume MahlerIssue in the makefile in release-8doing make clean failed when I am in release-8 branch
The lines which fail is
"""
...
checking for /Users/gmahler/soft/wcstools-3.8.4/libwcs/include/wcs.h... no
configure: error: Include file wcs.h not found
make: *** [config.status] ...doing make clean failed when I am in release-8 branch
The lines which fail is
"""
...
checking for /Users/gmahler/soft/wcstools-3.8.4/libwcs/include/wcs.h... no
configure: error: Include file wcs.h not found
make: *** [config.status] Error 1
"""
in origin master I do not have this issue because make does not go in include
Here is the line when I do make clean in origin master
"""
...
checking for /Users/gmahler/soft/wcstools-3.8.4/libwcs/wcs.h... yes
checking for pix2wcs in -lwcs... yes
...
"""Eric JulloEric Jullohttps://git-cral.univ-lyon1.fr/lenstool/lenstool/-/issues/10Release-8: broken new implementation of g_ampli_m3_boucle with wcslib in g_am...2023-06-02T15:28:55ZBenjamin BeauchesneRelease-8: broken new implementation of g_ampli_m3_boucle with wcslib in g_ampli.cI spotted two problems with the implementation of wcslib in the release-8 branch. A typo in keep_cl that I corrected and the new implementation of the function mentioned in the title. I solved the second one by returning to the previous ...I spotted two problems with the implementation of wcslib in the release-8 branch. A typo in keep_cl that I corrected and the new implementation of the function mentioned in the title. I solved the second one by returning to the previous code and changing the line where the wcstool function was called.
@ejullo as you made the modification to make it more efficient, so I left your code in the g_ampli.c but commented. Hence, we have something working now. I didn't find what was wrong; it does not work in single or multi-thread mode. I put in the attachment a simple model with the results that I got before my fix.
I open the issue if you want to rework your function @ejullo, but if you are ok with my fix, I will close the issue and remove the commented part. But I am not sure it is super robust with the MAX_IMAGES assumption.
[Ampli-problem.zip](/uploads/4571268a40f91e5438bcf97d179beb41/Ampli-problem.zip)Eric JulloEric Jullohttps://git-cral.univ-lyon1.fr/lenstool/lenstool/-/issues/5SIS profile does not produce radial critical curve2023-06-02T09:16:17ZGuillaume MahlerSIS profile does not produce radial critical curveThe SIS profile in lenstool only produce the external critical curve, !
This screenshot shows it on the left the magnification map at z_s=2 with the critical curve, on the right, it is the zonemult map created by lenstool which does not ...The SIS profile in lenstool only produce the external critical curve, !
This screenshot shows it on the left the magnification map at z_s=2 with the critical curve, on the right, it is the zonemult map created by lenstool which does not yield the correct results either.
[Screenshot_2022-11-05_at_15.03.00](/uploads/0970e978ad43fbd5d4422aa893c58149/Screenshot_2022-11-05_at_15.03.00.png)
and it should be something that looks this this
http://gravitationallensing.pbworks.com/f/1218447069/SIS_soft.png
I attached here a tar file containing the lenstool files and map:
in this directory launch SIS.par to reproduce what I show in the image.
[lenstool_issue_SIS_critic.tar.gz](/uploads/3d8410cbba3e3babbbb81157ca871d68/lenstool_issue_SIS_critic.tar.gz)Johan RichardJohan Richardhttps://git-cral.univ-lyon1.fr/lenstool/lenstool/-/issues/8Apparent problem with the lenstool multiplane and the potential 17 (external ...2023-02-24T16:09:25ZBenjamin BeauchesneApparent problem with the lenstool multiplane and the potential 17 (external map)As the potential 17 is defined by an external map and its WCS infos, it does not have a `lens[j].Cinit` parameter set. This is used to deflect the background lens according to the foreground one, so it should not behave properly in that ...As the potential 17 is defined by an external map and its WCS infos, it does not have a `lens[j].Cinit` parameter set. This is used to deflect the background lens according to the foreground one, so it should not behave properly in that context for now.Benjamin BeauchesneBenjamin Beauchesnehttps://git-cral.univ-lyon1.fr/lenstool/lenstool/-/issues/2Adding "flux 1" functionality in cleanlens while pixel 2 is active in runmode...2020-02-28T19:57:56ZGuillaume MahlerAdding "flux 1" functionality in cleanlens while pixel 2 is active in runmode (flux/pixel in source plan -> image plan)While using lenstool and pixel 2 to send an image from the source plan in the image.
cleanlens section do not handle the keyword:
flux 1
flux 1 was created to consider pixel value as flux/pixel and not flux/arcsec (i.e. surface brillian...While using lenstool and pixel 2 to send an image from the source plan in the image.
cleanlens section do not handle the keyword:
flux 1
flux 1 was created to consider pixel value as flux/pixel and not flux/arcsec (i.e. surface brilliance)
This functionality is present while using pixel 1.
Documentation does not mention this.
"""""""""""""
runmode
reference 3 3.586259 -30.400174
pixel 2 720 output_im.fits
end
cleanlens
cleanset 1 2
flux 1
sframe input_im.fits
end
"""""""""""""Johan RichardJohan Richardhttps://git-cral.univ-lyon1.fr/lenstool/lenstool/-/issues/1Lenstool .par Behaviour2019-07-19T07:18:17ZJohan RichardLenstool .par BehaviourI list here some of the behaviours from the .par file.
- The only sections which are mandatory are the "potential" section and the "fini" keyword
- If "grille" is NOT provided then everything works but no lensing is performed
- If "gr...I list here some of the behaviours from the .par file.
- The only sections which are mandatory are the "potential" section and the "fini" keyword
- If "grille" is NOT provided then everything works but no lensing is performed
- If "grille" is provided with nlens = 0 then nothing happens (only pot.dat is written empty)
=> In that case it should be the same as when grille is NOT provided: actions are performed without any lensing effect.
- If no "potential" section is provided the program stops with an error (no potential)
=> In that case it should simply be a warming and no lensing effect is applied.
- If no "champ" section is provided the default [-20,20;-20,20] region is used but no message is given
=> In that case it should write a warning that the default values are used (and mention them)
- If no "cosmology" section is provided the default cosmology is used but no message is given
=> In that case it should write a warning that the default values are used (and mention them)
- If the "fini" keyword is missing the program loops infinitely.
=> Lenstool should detect EOF in the .par and stop.Johan RichardJohan Richard