Comments for The AstroStat Slog

Comment on [Jobs] postdoc position at UC Berkeley by Joseph Hilbe

Joseph Hilbe — Wed, 27 Jan 2010 04:34:55 +0000

The web site for the new ISI astrostatistics committee and Network, which is aimed to become a global network of astrostatisticians is:
http://isi.cbs.nl/COMM/AstroStat/index.htm
Anyone with an interest in the statistical analysis of astronomical data is welcome to join. The current website is only the initial version and will be updated shortly.

Joseph Hilbe
Chair, ISI astrostatistics
hilbe@asu.edu

Comment on [MADS] Parallel Coordinates by Alfred Inselberg

Alfred Inselberg — Tue, 26 Jan 2010 12:56:32 +0000

I would like to announce that

Parallel Coordinates – This book is about visualization, systematically incorporating the fantastic human pattern recognition into the problem-solving …
http://www.springer.com/math/cse/book/978-0-387-21507-5

is now available. Among others, Stephen Hawking complimented the book and recommended that his students read it.

Comments are invited

Alfred Inselberg

Comment on AstroStatistics School in India by Joseph Hilbe

Joseph Hilbe — Mon, 11 Jan 2010 21:06:43 +0000

I invite all those with an interest in the statistical analysis of astronomical data to join the International Statistical Institute’s (ISI) Astrostatistics Network. Last month (December 2009) the ISI executive committee and Council approved the creation of a standing astrostatistics committee and Network, the first such committee authorized under a major statistical or astronomical organization. The web site and information about the committee and Network is: http://isi.cbs.nl/COMM/AstroStat/index.htm
Feel free to contact me directly about the Network, hilbe -at- asu.edu.
Joseph M Hilbe
Chair, ISI astrostatistics

Comment on People by nestor

nestor — Wed, 16 Dec 2009 11:13:10 +0000

Thanks for the comment Vinay. Yes I see that. I am just intrigued at the
way people bin and report their findings. A 0.2 sigma shift for a
couple of sources would create a nearly normal
distribution. This experiment has only detected the Crab nebula at 6 sigma
and now it’s making a significant leap in its detection limits albeit at a
purported 2 sigma. The equivalent in X-rays would be an experiment that only
detects is the Crab at 5 sigma, sees nothing in between and then
publishes a paper claiming two sigma detections for the faintest sources
in the Chandra deep field under the same argument. Would you believe
such a claim?. It looks statistically sound but there is counterintuitive here
that just doesn’t compute. Thanks again.

Comment on Do people use Fortran? by hlee

hlee — Wed, 09 Dec 2009 21:20:39 +0000

Thanks for sharing your experience with us. I heard of TCL a few times. I’m glad that I didn’t dig into it because I don’t like nightmares. But knowing baseline languages seems important for me to understand how analysis is done, to dissect the analysis procedure for robustness, and to measure the reliability of quantification. Frequently, from data scientist perspective, I find that analysis packages are lack of algorithmic explanation, particularly when it comes to statistical inference where error propagation and dependency cannot be presented by one single sigma.

Comment on Do people use Fortran? by Elbereth

Elbereth — Mon, 07 Dec 2009 17:20:13 +0000

I have been an undergraduate in Spain, and I can assure you we are still taught Fortran (in fact it’s considered somewhat innovative, which is a bit surprising when you realise most of the teachers only know Fortran 77). I was lucky enough to get the only teacher that did work with Fortran 90 (The gotos give me the creeps, and the pseudo-pointers of Fortran 95 are even more scary).
My experience in astronomy is brief, I’m a PhD student, therefore my knowledge of the subject is limited, but as far as I know that people working in high energies (X-rays and probably some people working in Gamma-rays too) used to rely on Fortran until very recently, due to the main analysis tool being XSPEC v11 (and previous versions), all based on Fortran, I believe. The newer XSPEC versions (v12+) are based on TCL (which I personally consider a nightmare of a language), and C/C++.
So we are not really encouraged to learn Fortran nowadays, unless we want to work on modelling or MHD simulations, but it always comes in handy, if only to understand some of our tutors’ quirks so we end up learning it anyway. The same could probably apply to C.

Comment on People by vlk

vlk — Fri, 04 Dec 2009 17:30:36 +0000

It seems that what they are doing is to compute a significance for every source in their catalog and compare the distribution of significances to what would be expected from a Gaussian distribution. They find 7 sources at significances >2sigma. Seeing 7 out of 27 (25%) at >2sigma is not very probable, so, in conjunction with the coincidence of these sources with known pulsars, they claim the detection. 2 sigma implies that 5% of the time a random deviation can produce numbers greater than that threshold. Because it is a two-sided distribution, on the upper side that is a 2.5% probability, which, for 27 sources, implies 0.6 sources are expected to exceed that number.

I think the methodology is fine. It is a combination of marginal detections superposed on known catalog sources that makes the cut.

Comment on People by nestor

nestor — Fri, 04 Dec 2009 09:03:56 +0000

has anyone read this paper?. it has been accepted by ApJ Letters.

http://arxiv.org/abs/0912.0386

can multiple 2-sigma detections a detection make without stacking?. i would appreciate any comments. a baffled scientist.

Thanks,
nestor

Comment on some python modules by hlee

hlee — Thu, 19 Nov 2009 21:10:35 +0000

I appreciate your kind words and my apologies for the inconvenience by RSS feed failure. From time to time, I look for a solution but most of description assumes a certain directory structure for the wordpress blogging package. This slog not being located under my home directory makes a fix difficult. Even if live feed fails from time to time, it generally has worked out after correcting the rss feed url. To answer your question, as mentioned, some motivation helps learning a new language. It was not python that I was banned from but a data analysis package that uses python. Not being allowed to use that package, I lost a reason to use python, while most of my stuffs can be done with R and a bunch of others (IDL,c, Fortran,c-shell script, pearl, Matlab, Mathematica, SAS, Minitab, SPSS, and few more). The ability of communicating and embracing other languages is the great merit of python, I think. I'm happy to know that you find these modules and libraries are informative. Thanks!

Comment on some python modules by robert

robert — Wed, 18 Nov 2009 09:38:48 +0000

Thanks for the useful links! I’m in the process of switching from IDL to python and some of these modules are new to me. Your opening sentence makes me a bit worried though! I was really torn between R and python, but eventually considering also Rpy I went with python. What are your main arguments against python? I’m really curious, whether I should reconsider…

Also – I was following this blog from the beginning, but roughly a year ago it seemingly died. I was suprised to find many new posts here today when I actually visited the website. Could there be a problem with your RSS feed? Neither my google reader nor NetNewsWire see posts after 11th Dec 2008.

Thanks and keep up the nice blog!