April 2002 - Shape-univ - cosmo.torun.pl

DE-V0.01 compilable + runnable -> antipode ?
by Boud Roukema 13 May '02

13 May '02

OK everyone, Thanks to your comments, DE-V0.01 has now been released! There's not yet any real documentation on the program modules themselves, but what I've done is documentation on how to manage the whole package. Compiling different routines and linking them together is generally done with a makefile, but different versions of linux/unix use different syntax or levels of generality, so I've followed the pgplot solution which is to have a shellscript file makemake which makes the makefile depending on what system you're on. Well, if you download the new package (either with or without the data files; "nodat" means without the data files) from: http://www.astro.uni.torun.pl/~boud/DE_tar/ and then read through the web pages: http://www.astro.uni.torun.pl/~boud/DE/ you should be able to make the library (libDE.a) and compile a program like antipode and see at what redshift the antipodes are. In fact, if you just type make antipode after you've made the makefile, then libDE.a should be made for you automatically... Next on my list: - documentation within each of the more important routines - cosmological routines - routines for reading data - main programs which do the work Cze�� Boud

1 1

off-topic: taxing free software - total nonsence!!!!!!
by Michal Frackowiak 26 Apr '02

26 Apr '02

Polish taxing institutions were always hungry for taxing free software. But so far they only tried to tax USERS. Now they are preparing to tax PROGRAMMERS!!!!!!!!!!!!!!!!!!!!!!!!!!! So far the new law gives the chance to tax programmers but the question is if the taxmen will be really so ignorant about the idea of free software? Unofficially the Minister of Finances is supporting the idea of taxing free software. But we shall wait..... The paradox is because the novelisation of tax law assumes giving a licence (selling) should be taxed. And that is wise. Free programmes also give licences (mainly GPL). On the other hand if a given service is done free of charge when one usually get paid for it, this service should be taxed. => giving a FREE licence is not common, so it is an exception, because licences are always SOLD. free programers give for free what is usuallu sold. => they should be taxed It is sad. No more comments. Michal F. http://www.gazetaprawna.pl/numery/05602-661/aktualnosci.html?id_gpol2=3554

1 0

Intro to cosmology (my master thesis) %LOOK%
by Michal Frackowiak 26 Apr '02

26 Apr '02

Hej! Ponizej posylam link do szkicow mojej pracy magazyniersiej. Mam nadzieje, ze moze to posluzyc do "lagodnego" wprowadzenia do kosmologii. To, co sie tam znajduje w tej chwili, to: * wstep (historyczny?) i kilka slow o kosmologii, problem stalej kosmologcznej * podstawy teoretyczne: r-nia Einsteina, dodatkowe pole kwintesencji - z metod wariacyjnych, wyjasienia podstawowych pojec w kosmologii * troche szczegolow o samej kwintesencji * i wiele wiecej, ale to pozniej (supernowe Ia, anizotropia CMB) prosze nie zwazac na literowki i bledy, tekst jest napisany, ale go jeszcze nie czytalem ;) mam nadziej, ze bedzie uzyteczny pozdrawiam Michal http://www.ncac.torun.pl/~michalf/MasterThesis/

1 0

DEflow1.eps printable version
by Boud Roukema 26 Apr '02

26 Apr '02

Cze�� The flowchart I sent yesterday was written with xfig There is a bug in xfig when writing a postscript file - it does not write the final end statement showpage So you can see the file with ghostview or gv, but if you print it nothing will happen - sorry. The solution to this bug is to edit the postscript file, and add the line showpage as the last line. Then the file is printable. Here is a printable version of DEflow1.eps. Boud

1 0

rough flowchart
by Boud Roukema 25 Apr '02

25 Apr '02

Cze�� wszystkim, I've prepared a flowchart and realised that only *one* of the "cosmological" function/routine files - pdWeiq.f - is called by plates, (pdWeiq or pdWein is called for getting radial proper distances), but angle4D8 (in angle4D) which I've called "numerical" is also called and is actually very useful conceptually - it's what is used to get the distance separation between two points in arbitrarily curved space. [Just as a "distance" on a 2-sphere is an arc-length on a 2-sphere embedded in R^3, a "distance" on a hypersphere or 3-hyperboloid is an "arc-length" on a 3-sphere or 3-hyperboloid embbeded in R^4 or M^4 (Minkowski-4).] * the flowchart will be correct once I modify the names to add "DE" and add documentation, and I haven't yet written DErcurvq (which should be like rcurvfn in pdWein.f) * the arrows show function calls. Na jutro pi�tek o 14:00! Boud

1 0

documentation; What name for our package?
by Boud Roukema 17 Apr '02

17 Apr '02

Cze�� wszystkim, I am very glad that several of you (at least three people) have had the courage to tell me that I have to do some work (in documenting the software), otherwise we will not get anywhere in our "monograph course"! This a good sign for learning how to work together :). So, I'm about to start preparing some documentation. My plan is to follow the PGPLOT style documentation which was used for the ArFus galaxy formation package: http://darc3.obspm.fr/~roukema/ArFus/index.html (You can download a *local* copy of the full package, including small example N-body simulations here - total size 31Mb: http://www.astro.uni.torun.pl/~boud/ArFus_tar/ArFus-V0.03.tar.gz ) Question: what name do we want to call our package? I think the name should not be too long, and it would be good to have a two-letter abbreviation to put in the subroutine (module) names, like "PG" for pgplot, "AF" for ArFus. Some ideas with comments + for - against ... ShapeUniv - SU + easy to say - confusion with SU groups of superstring theory? - confusion with Soviet Union? DarkEnergy - DE + easy to say - confusion with the name of a neighbouring country? FormaWszech�wiata - FW - confusion with "forward" in emails - difficult to say "FW..." for the name of a subroutine - includes a non-latin-1 character ObservCosm - OC - pretentious (claims to cover all of observational cosmology) TorunCosm - TC + maybe the best? - confusion with TCP/IP? QuintEssence - QE - confusion with Queen Elizabeth? CosmParameters - CP --- confusion with Communist Party KoperNik - KN - too many things are called Copernicus in Toru� & Poland - pretentious? Well, I've spent enough time on this, I'm sure others have some ideas. I'll put "DE" for the moment - we can change this later with a shell script depending on what people choose. Na ra� Boud

2 1

summary 14/04/2002
by Boud Roukema 12 Apr '02

12 Apr '02

Cze�� wszystkim, In today's lecture we regrouped, the two groups were Rafa�/Jacek/Hubert/Micha�-F Marcin/Pawe�/Boud In my group, Marcin and Pawe� looked through some of the subroutines in cosm.tar.gz and nrec.tar.gz and I tried to explain what the important ones did, and in the other group people tried compiling. * documentation It is clear that I have to write some good documentation, and quickly, so that everybody is not slowed down by illegible programs... Rafa�'s suggestion for documentation: - what the program/subroutine does - what external program calls there are - what data it uses * reference books Other discussion was about reference books for modern cosmology, some of the good recent books are: - Peebles - Liddle - Peacock Maybe someone could find out if we have these in our library, or who has borrowed the copy(ies) of these. Another reference book we should have is Numerical Recipes. Probably later when we have some grant money, we could follow IUCAA/IAP type policy and have something like 2-3 copies of each of these key references. (But this also requires trusting each other - the books are expensive and need to be shared around and kept track of by the library!) Of course, online references cost nothing, though of course it's nice to have printed books. I think online sources are probably the best for correlation functions and w_q. * computer access Andrzej M - could you coordinate with Hupert to tell him about what modem access we have, if any? He lives in a small town 30km from Toru� and apparently has no local facilities except his own PC. OK, watch this space for documentation, let's see if I can get organised... Pozdrawiam boud

1 0

... we're in pracownia computerowa w radio astronomii i w biurze bouda...
by Boud Roukema 12 Apr '02

12 Apr '02

1 0

14:00-15:00 cosm routines + antipodes; 15:00-15:30 feedback
by Boud Roukema 12 Apr '02

12 Apr '02

Cze�� wszystkim, Sorry for writing so late - it just goes to prove that my own organising skills need improving! Last week we more or less went through the points I had suggested. I thought that dividing into small groups with a mix of "generations" was best; Micha� F thought that people who already know each other was best. The final result was to divide people randomly, so here are the initial groups: Tomek/Micha�-F/Rafa� Radek/Hubert Marcin/Andrzej-M/Boud We have at least one person with a computer in an office at Piwnice, so the plan was to have the lecture as follows: * lecture structure Friday 14:00-15:00 each group meets around a computer and see what can be done for playing with the programmes, learning how they work and working towards the correlation function etc programmes 15:00-15:30 we meet back together in the lecture room, have coffee and discuss how things went, problems, what to do next, etc. (people who do not catch the bus can stay later) * comment: some people have more time, some have less time (e.g. only the Friday lecture time). I don't see this as a problem - different people should be able to accept complementary time usages. This was part of the exercise with getting to know one another... [English language note: "complement" .ne. "compliment"] * I still think this is a good plan - I got some feedback that Rafa� and Micha� could not compile anything of my programs ;), but below is some initial stuff which I think should be able to keep people occupied for 1 hour. If you can get through that, you'll have some idea of my programming style and how to decode it ;) or improve on it. Exercise: find out some local cosm parameter values for which we could see the Galaxy (or the spot where the Galaxy was going to form), for a trivial topology! ###################################################################### **** Today, 14:00 12 kwi�tna 2002 **** ----- I will wait at the lecture room for 10-15 minutes (14:00-14:15) for any new people that turn up, or people that were missing last week. E.g. Pawel, Bartek, Micha�-H, if you arrive I'll explain to you what we did last week and you can join the existing groups (or if there are too many people, we can start a new group). ---- ---- At 15:00 we meet back at the lecture room and give/receive feedback. ---- ###################################################################### See you in 20 minutes! Micha�, thanks for the C references! Boud To get people started, here is a rough guide to cosm.tar.gz ---------------------------------------------------------------------- cosm.tar.gz e.g. unload in a directory called cosm/ ## x makefile, 1681 bytes, 4 tape blocks - you'll need some routines in nrec.tar.gz, e.g. unload in a directory called nrec/ - you should read through the makefile in nrec.tar.gz, make some directories, and then type make nrec.a in the directory with the source and then make appropriate links back in the cosm/ directory ## x d_l3.f, 14623 bytes, 29 tape blocks ** proper motion distance ** - callable functions - calculates proper motion distance as a function of z - calculates inverse function: z as a function of proper motion distance - integrates an initial table the first time it's used - test program at beginning - depends on Om_m, Om_Lambda (called "cc_0"), w_q test compile: switch "program" and "subroutine" at top, and make d_l3 x d_l.f, 732 bytes, 2 tape blocks - like d_l3.f but no Lambda, w_q dependence x d_l2.f, 14135 bytes, 28 tape blocks - like d_l3.f but no w_q dependence ## x pdWeiq.f, 4968 bytes, 10 tape blocks ** proper distance ** - callable functions - calculates proper distance as a function of z - calculates inverse function: z as a function of proper distance - integrates an initial table the first time it's used - test program at beginning - depends on Om_m, Om_Lambda (called "cc_0"), w_q test compile: switch "program" and "subroutine" at top, and make pdWeiq x pdWein.f, 6835 bytes, 14 tape blocks - like pdWeiq, but no w_q dependence ## x tofz2.f, 6642 bytes, 13 tape blocks - callable functions: cosmological time as a function of z and inverse - no allowance for (w_q.ne.-1) test compile: switch "program" and "subroutine" at top, and make tofz2 ## x antipode.f, 16710 bytes, 33 tape blocks - This should be a fun test of your geometry intuition. Depending on values of Om_m, Om_Lambda, w_q, find out at what redshifts the "South Pole" and "North Pole" are (let's say that we live at the North Pole, after all, pogoda tutaj jest prawie blisko z pogodzem z Polem P�nocnej... ;) test compile: switch "program" and "subroutine" at top, and make antipode ## Exercise: find out the most "reasonable" Om_m, Om_Lambda, w_q values for which the "South Pole" etc. are at observable redshifts. ## Other stuff, probably less useful for our project. x cdm.f, 1910 bytes, 4 tape blocks x dVdz.f, 3559 bytes, 7 tape blocks x dVdz2.f, 4931 bytes, 10 tape blocks x schech.f, 1806 bytes, 4 tape blocks x units.f, 157 bytes, 1 tape blocks x zt.f, 9294 bytes, 19 tape blocks x zt_noL.f, 8746 bytes, 18 tape blocks ----------------------------------------------------------------------

1 0

programs + data
by Boud Roukema 12 Apr '02

12 Apr '02

Hi, Here are my existing programmes + data. There are certainly a few things like filenames and logical options which need to be set correctly for these to work - they are not (yet) user friendly. But you can see (more or less) what I've done. Boud

2 1