- IMP-dev - salilab.org

IMP 2.1
by Daniel Russel 01 Oct '13

01 Oct '13

It is time to start on the 2.1 release. There are currently two issues open in the tracker attached to that milestone - one regarding isd (#526) - and the MC mover/optimized behavior. I'd like to resolve the latter by adding checks so errors are thrown with most movers if they try to move non-optimized attributes, which should be a relatively quick change that I can do this afternoon. Otherwise, is there anything else that someone thinks should be cleaned up before the release?

3 3

set_coordinates_are_optimized function
by Davide Baù 27 Sep '13

27 Sep '13

Hi all, I have a question about the "set_coordinates_are_optimized" function in IMP 2.0.1. In the IMP versions < 2.0, to set the coordinates of a given particle, e.g., P1 to (0,0,0), I would use the following code: > fixedP = "1" > p = ps.get_particle(index) > p.set_name(pname) > if pname == fixedP: > d = IMP.core.XYZ(p) > d.set_coordinates(IMP.algebra.Vector3D(0,0,0)) > d.set_coordinates_are_optimized(False) In IMP 2.0.1, even though, at this point, the coordinates of the particle "fixedP" are set to (0,0,0), the "set_coordinates_are_optimized" does not seem to work anymore (the coordinates of "fixedP" change during the optimization). Has this function beed replaced with a new one in the latest IMP version? Thanks, Davide

5 15

IMP 2.1 release
by Daniel Russel 21 Sep '13

21 Sep '13

The 2.1 release is supposed to occur in a week or so. Due to various things keeping me busy at the moment (eg the end of my time at Google), I propose we delay it by a couple of weeks until say October 18. Any objections?

3 2

IMP 2.1 release
by Daniel Russel 23 Aug '13

23 Aug '13

The scheduled date for the IMP 2.1 release is coming up in about a month. If there is anything that it is important happen before then, please make there is an issue in the tracker that is tagged with the 2.1 release milestone. Also, please check the tracker and see if there are issues assigned to you that are tagged for the release and deal with them (in the worst case by just marking them as for the 2.2 release). In addition, if you have many any particularly interesting changes to code, add a brief description to the ChangeLog.md. Some time after the 2.1 release, I'll start removing things marked as deprecated, so please clean up such usages in your code.

1 0

Problem compiling an old IMP version cloned from GIT
by Davide Baù 06 Jul '13

06 Jul '13

Hi all, I was trying to install an old version of IMP (12787) on my machine to do some tests. I tried downloading it using git (this should be the 12787 version): > git clone https://github.com/salilab/imp > cd imp > git reset --hard 847e65d44da7d06718bcad366b09264c818752d5 By doing so, the compilation fails with an error related to the dope_score.lib file: > Errors building: > build/data/atom/dope_score.lib It seems that the error arises because the two files dope_score.lib and dope_scorehr.lib are not anymore in the modules/atom/data dir. If I copy them from an old IMP directory, the compilation end with no problems. Everything works fine also using svn: > $ svn co http://svn.salilab.org/imp/trunk imp-r12787 -r 12787 where the dope_score files are still present: > imp-r12787/modules/atom/data$ ls -1 > SConscript > dope_score.lib > dope_scorehr.lib > par.lib > protein_ligand_pose_score.lib > protein_ligand_rank_score.lib > radii.lib > top.lib > top_heav.lib Have the dope_score files been removed in the new IMP versions? Thanks and best regards, Davide

2 5

For those of you who write documentation
by Daniel Russel 28 May '13

28 May '13

The number of warnings produced when running doxygen had gotten kind of out of control, making it quite hard to work on the documentation. To address this, various people have been assigned issues to fix their documentation bugs. To make this easier, I've refactored the documentation generation code to make it faster to check on warnings in a particular module. You can now do: - IMP-doc to build the documentation for all of IMP - IMP.core-doc to (re-)build the documentation for just IMP.core (remove doc/html/core/index.html if it has already been generated) This means, that once you make some changes to the documentation of your code, you can run a relatively quick build to make sure nothing broke/look at the result. See the change log (or run "git imp update" for more info). As a side note, if you are in the habit of documenting parameters in doxygen using the \param (or @param) commands, please be very careful. Getting a parameter name wrong, or leaving out a parameter results in a warning, and there are tons of these. This both makes the resulting documentation not very useful and makes it hard for anyone to work on the documentation. And one last note, you are, in general, best off using the latest version of doxygen as important bugs are fixed useful capabilities are added at a pretty rapid clip. It is quite easy to install by downloading from http://ftp.stack.nl/pub/users/dimitri/doxygen-1.8.4.linux.bin.tar.gz (on linux) or via homebrew on a mac. I would like to move towards treating doxygen warnings as errors so as to try to keep it warning free, but we are a ways from that. Thanks.

1 0

git flow -> git imp
by Daniel Russel 09 May '13

09 May '13

I found dealing with the README.md and other required was resulting in me using feature branches less than I would like so I forked git-flow and made an imp-customized version. It is pulled in by the the imp repository so you don't have to install anything. Next time you update imp, rerun setup_git.py and you will have several new commands available: - git imp update: updates the repository and displays any new changes listed in ChangeLog.md - git imp feature start: starts a new feature. It opens an editor (using $EDITOR) to create a README.md and commits it to the new feature branch. - git imp feature finish: finishes a feature. The README.md is used as the commit message for the feature and the README.md file itself is removed. The changes in the branch are rebased against develop as part of the process and then squashed into one commit. let me know if you find the scripts useful and if you encounter any problems.

2 2

Re: [IMP-dev] [imp] problem with euler angles (#303)
by Barak.raveh 06 May '13

06 May '13

Does anyone use Euler angles? The related functions don't work too well (numerical instability) so let us know if you need them / feel like fixing them / ok with getting rid of them / marking then as dangerous On Apr 24, 2013, at 2:53 PM, drussel <notifications(a)github.com> wrote: > Should probably just remove the functions. They are numerically unstable and no one has any interest in fixing them. > > — > Reply to this email directly or view it on GitHub.

5 8

Easygit
by Daniel Russel 03 May '13

03 May '13

I just stumbled upon easygit. It might be something for people who are intimidated by git to try out: <http://people.gnome.org/~newren/eg/git-for-svn-users.html> homebrew includes it.

1 0

imp doc
by Yannick SPILL 30 Apr '13

30 Apr '13

Hi there Which part of the imp docs (doxygen+github) can we consider up-to-date? Which part is outdated, or refers to deprecated functionality? thanks Y

2 1