From boud w astro.uni.torun.pl Fri Apr 5 01:06:25 2002 From: boud w astro.uni.torun.pl (Boud Roukema) Date: Fri, 5 Apr 2002 01:06:25 +0200 (MET DST) Subject: wyklad 5 kwietnia 2002 Message-ID: Cze�� wszystkim, The following is my suggested agenda for the lecture this afternoon 14:00 Fri 5 Apr 2002 in the Radio building. (1) First things to play with :) ----------------------------- cosm.tar.gz nrec.tar.gz You'll find these attached - I haven't checked them, sorry, I'm tired and it's 01:06. (2) Poll (for finding out how to work together) -------- During the lecture I will ask everyone (including me!) to rate how he/she feels his/her level is from -5 (low) to +5 (high). The purpose is so that we each get to know where each other feels strong and/or weak, and we can help each other in their weak points. * self-confidence * organising skills * unix * fortran * C * pgplot * typing speed * comoving coordinates * curvature * topology * correlation function * null hypothesis rejection I will suggest splitting up into pairs/triplets and discussing these in separate groups for 20 minutes (that's less than 2 minutes per question, so you may not be able to discuss everything). Within each pair/triplet, please look at the strengths/weaknesses of the other 1 or 2 people and decide how you can concretely help the others in their weaknesses. In particular, for self-confidence, it would be good over the next week to make at least 1 compliment about the other 1 or 2 people in the pair/triplet. It should be a genuine compliment - and this will force you to try to really understand the person, otherwise it can't be genuine... (3) time scale -------------- 8 weeks till 31 May (4) possible programme ---------------------- a - play with cosm.a + nrec.a and document the routines in English & po polsku, see how much can be thrown away because redundant and/or useless and/or confusing b - get everything compiled including b1 -- reading in data files b2 -- calculating correlation functions b3 -- plotting correlation functions b4 -- plotting null hypothesis rejection levels c + add new routine for modelling ubr selection and avoiding z scrambling d + can calculation be speeded up? e + calculate for full 11K initial release f + write routine for reading in Las Campanas Distant Cluster Survey http://de.arXiv.org/abs/astro-ph/0201534 g + calculate result (should have degeneracy similar to SNe Ia degeneracy) h - write down method and results in English i ** positive/negative feedback loop ** Think about the results and based on thoughts, corrections, worries, iterate again through loop doing it better. j Keep iterating until convergence is achieved. kwiecie� 2002 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 maj 2002 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 (5) Question: ------------ Is it reasonable to ask everyone to write at least one email per week to the list saying where they've got to, what their problems, successes are? (6) Conceptual task: ------------------- getting null hypothesis rejection probabilities for the "D" test; how reliable, robust is the test? Useful reference? (at least for one of Boud's weaknesses, see (2) above ;) ) ----------------- http://www.le.ac.uk/cc/iss/tutorials/cprog/cccc.html Na popo�udniu o 14:00 (not 14:15)! Boud -------------- następna część --------- Binarny załącznik wiadomości został usunięty... Nazwa: cosm.tar.gz Typ: application/octet-stream Rozmiar: 17057 bytes Opis: cosm.tar.gz Adres: -------------- następna część --------- Binarny załącznik wiadomości został usunięty... Nazwa: nrec.tar.gz Typ: application/octet-stream Rozmiar: 35657 bytes Opis: nrec.tar.gz Adres: From boud w astro.uni.torun.pl Thu Apr 25 17:42:50 2002 From: boud w astro.uni.torun.pl (Boud Roukema) Date: Thu, 25 Apr 2002 17:42:50 +0200 (MET DST) Subject: rough flowchart Message-ID: 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 -------------- następna część --------- Binarny załącznik wiadomości został usunięty... Nazwa: DEflow1.eps Typ: application/postscript Rozmiar: 8063 bytes Opis: rough flowchart Adres: From boud w astro.uni.torun.pl Fri Apr 26 16:44:12 2002 From: boud w astro.uni.torun.pl (Boud Roukema) Date: Fri, 26 Apr 2002 16:44:12 +0200 (MET DST) Subject: DEflow1.eps printable version Message-ID: 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 -------------- następna część --------- Binarny załącznik wiadomości został usunięty... Nazwa: DEflow1.eps Typ: application/postscript Rozmiar: 8072 bytes Opis: printable version Adres: From michalf w ncac.torun.pl Fri Apr 26 19:49:22 2002 From: michalf w ncac.torun.pl (Michal Frackowiak) Date: Fri, 26 Apr 2002 19:49:22 +0200 (CEST) Subject: Intro to cosmology (my master thesis) %LOOK% Message-ID: 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/ From michalf w ncac.torun.pl Fri Apr 26 20:04:21 2002 From: michalf w ncac.torun.pl (Michal Frackowiak) Date: Fri, 26 Apr 2002 20:04:21 +0200 (CEST) Subject: off-topic: taxing free software - total nonsence!!!!!! Message-ID: 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 From boud w astro.uni.torun.pl Fri Apr 5 13:02:18 2002 From: boud w astro.uni.torun.pl (Boud Roukema) Date: Fri, 5 Apr 2002 13:02:18 +0200 (MET DST) Subject: programs + data Message-ID: 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 -------------- następna część --------- Binarny załącznik wiadomości został usunięty... Nazwa: 2dF.tar.gz Typ: application/octet-stream Rozmiar: 457967 bytes Opis: shape-univ project: src + inputdata Adres: From michalf w dwarf.ncac.torun.pl Fri Apr 12 12:35:13 2002 From: michalf w dwarf.ncac.torun.pl (Michal Frackowiak) Date: Fri, 12 Apr 2002 12:35:13 +0200 (CEST) Subject: C programming guides In-Reply-To: Message-ID: Here are some guides I have found: http://www.gnu.org/manual/glibc-2.2.3/ - the best one - complete desc. of linux's C standard library http://www.ulib.org/webRoot/Books/Numerical_Recipes/ - a book with numerical codes, also in C (F, F90 and C++ as well) http://www.infosys.utas.edu.au/info/documentation/C/CStdLib.html - ANSI standard library - HARDCORE!!! - but necesarry There were some useful guides but now seem abandomed - perhaps I have them at my home comp. You can also do a 'info gcc' to have a description of the compiler and the language - quite useful! You might also want to search with www.google.com - sure there are lot of c-guides. regards - Michal From boud w astro.uni.torun.pl Fri Apr 12 13:36:31 2002 From: boud w astro.uni.torun.pl (Boud Roukema) Date: Fri, 12 Apr 2002 13:36:31 +0200 (MET DST) Subject: 14:00-15:00 cosm routines + antipodes; 15:00-15:30 feedback Message-ID: 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 ---------------------------------------------------------------------- From boud w astro.uni.torun.pl Fri Apr 12 14:18:05 2002 From: boud w astro.uni.torun.pl (Boud Roukema) Date: Fri, 12 Apr 2002 14:18:05 +0200 (MET DST) Subject: ... we're in pracownia computerowa w radio astronomii i w biurze bouda... Message-ID: From boud w astro.uni.torun.pl Fri Apr 12 16:30:06 2002 From: boud w astro.uni.torun.pl (Boud Roukema) Date: Fri, 12 Apr 2002 16:30:06 +0200 (MET DST) Subject: summary 14/04/2002 Message-ID: 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 From boud w astro.uni.torun.pl Tue Apr 16 16:00:19 2002 From: boud w astro.uni.torun.pl (Boud Roukema) Date: Tue, 16 Apr 2002 16:00:19 +0200 (MET DST) Subject: documentation; What name for our package? Message-ID: 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 From amr w astro.uni.torun.pl Wed Apr 17 10:18:42 2002 From: amr w astro.uni.torun.pl (Andrzej Marecki) Date: Wed, 17 Apr 2002 10:18:42 +0200 (MET DST) Subject: documentation; What name for our package? In-Reply-To: from Boud Roukema at "Apr 16, 2002 04:00:19 pm" Message-ID: <200204170818.KAA03905@galileo.astro.uni.torun.pl> > ShapeUniv - SU > + easy to say Agree! > - confusion with SU groups of superstring theory? Maybe... > - confusion with Soviet Union? Well, maybe. But since *that* SU has gone and we all know it has, then it's no longer a problem, I think. > DarkEnergy - DE > + easy to say Agree! > - confusion with the name of a neighbouring country? I see NO problem here! Now, how about merging these two? We have a few options here: 1. DESU. Again it's easy to say but (depending on the pronounciation adopted) it may sound like a French word. ;-) 2. DEUS (dark energy + Universe's shape). Hmmm, "deus" is a latin word which may suggest we are... theologists. 3. SUDE. Better now? [...] > KoperNik - KN No, please... > - too many things are called Copernicus in Toruń & Poland YES!! (In Torun roughly a dozen.) > - pretentious? YES!!! > 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. I also vote for "DE". -- Andrzej From boud w astro.uni.torun.pl Thu Apr 18 02:41:59 2002 From: boud w astro.uni.torun.pl (Boud Roukema) Date: Thu, 18 Apr 2002 02:41:59 +0200 (MET DST) Subject: DE-V0.01 compilable + runnable -> antipode ? Message-ID: 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