Shape-univ

shape-univ@cosmo.torun.pl

176 discussions

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

wyklad 5 kwietnia 2002
by Boud Roukema 04 Apr '02

04 Apr '02

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

1 0

holiday 29 march
by Boud Roukema 28 Mar '02

28 Mar '02

Noone has said anything about whether or not they want a lecture tomorrow 29 March, but since it is officially a University holiday, I'll assume by default that noone expects a lecture. But if anyone wants to come along and talk, feel free to do so! I'll be here (at Piwnice). Cze�� Boud

1 0

feedback 15 March lecture?
by Boud Roukema 18 Mar '02

18 Mar '02

Cze�� wszysce (1) starting time 14:00 proposed In order that people can leave at 15:35 to catch the 15:52 bus, Andrzej proposed to start the sessions at 14:00 instead of 14:15. Everybody present was in favour, I think. In case lunch arrives late at Piwnice, we could start the lecture informally in the dining room - I see no problem with this, it could help to get people relaxed and talking informally. So, could people please be ready to start at 14:00, in the interests of 15:52-bus-catching people? Thanks. (2) Fri 15 March lecture I'm not sure if we went through enough or too much, but I realised that although there is nothing *difficult* in http://de.arXiv.org/abs/astro-ph/0106135, there are *many* details to think about. Rafa� asked about (i) what needs to be done to my programs, (ii) why can't we just use them unchanged? (ii) If we use them unchanged, then we will just get the old results, instead of exciting new results. (i) Here's a rough guess at what needs to be done: - rewrite the programmes to be more modular + add documentation - add a new routine for reproducing the photometric u-b_J-r band selection effects, so that "z-scrambling" doesn't have to be used for the "random simulations" for correlation function calculations - check that everything works fine on a subset and then calculate for the full "10K" data set on a fast computer - write input routines for the filter match selected candidate clusters in the 0.5 < z < 1 (approx) range (I'll look up the reference...) - this should directly reproduce the Omega_m - Omega_Lambda degeneracy of the SNeIa data (3) correlation functions Rafa� also asked about correlation functions. I suggest the 0106135 paper for the full details, but for an easy introduction, the following paper is probably better: http://cdsads.u-strasbg.fr/cgi-bin/nph-bib_query?bibcode=1994A%26A...285..3… So for next week (14:00 pi�tek 22 marca) I'll try to prepare some information about my programs and the programs themselves... Pozdrawiam Boud

1 0

Jump to page:

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

2004

2003

2002

Shape-univ