Comments for The AstroStat Slog http://hea-www.harvard.edu/AstroStat/slog Weaving together Astronomy+Statistics+Computer Science+Engineering+Intrumentation, far beyond the growing borders Fri, 01 Jun 2012 18:47:52 +0000 hourly 1 http://wordpress.org/?v=3.4 Comment on Wrong Priors? by Brendon J. Brewer http://hea-www.harvard.edu/AstroStat/slog/2007/wrong-priors/comment-page-1/#comment-1069 Brendon J. Brewer Fri, 01 Jun 2012 18:47:52 +0000 http://hea-www.harvard.edu/AstroStat/slog/2007/wrong-priors/#comment-1069 Hi, This is a very old post, but I just figured out what's going on. Carlos got wrong answers in his MCMC because with the diffuse prior there is a massive phase transition and you need Nested Sampling. There is nothing "wrong" about the diffuse prior in this problem. Cheers, Brendon Hi,

This is a very old post, but I just figured out what’s going on. Carlos got wrong answers in his MCMC because with the diffuse prior there is a massive phase transition and you need Nested Sampling. There is nothing “wrong” about the diffuse prior in this problem.

Cheers,
Brendon

]]>
Comment on From Terence’s stuff: You want proof? by Raoul LePage http://hea-www.harvard.edu/AstroStat/slog/2009/from-terences-stuff-you-want-proof/comment-page-1/#comment-1046 Raoul LePage Fri, 12 Aug 2011 17:44:19 +0000 http://hea-www.harvard.edu/AstroStat/slog/?p=4133#comment-1046 I've long been cautious about any model employing hypotheses of randomness in model components, particularly the backbone method of multiple linear regression, popular accounts of which may have encouraged the use of such hypotheses elsewhere. One particular liability of models is they may suggest particular classes of sample points are all that are needed. Examples include regression models positing iid errors that are much the same no matter which sample points are included in the regression design. Something similar is seen in time series and elsewhere. If the model is wrong on that point it may have encouraged a laxity in data selection from which there is no recovery once the data are in. I call this "brittle" modeling. Statistics is yet young. Stay tuned for models that are more in the spirit of sampling theory where inference is rooted in randomness introduced by the experimenter rather than hypothesized (a different sample space) and no model is correct although some may be estimated to provide better population description than others. Simpler may for some purposes prove better in the long run. I’ve long been cautious about any model employing hypotheses of randomness in model components, particularly the backbone method of multiple linear regression, popular accounts of which may have encouraged the use of such hypotheses elsewhere. One particular liability of models is they may suggest particular classes of sample points are all that are needed. Examples include regression models positing iid errors that are much the same no matter which sample points are included in the regression design. Something similar is seen in time series and elsewhere. If the model is wrong on that point it may have encouraged a laxity in data selection from which there is no recovery once the data are in. I call this “brittle” modeling. Statistics is yet young. Stay tuned for models that are more in the spirit of sampling theory where inference is rooted in randomness introduced by the experimenter rather than hypothesized (a different sample space) and no model is correct although some may be estimated to provide better population description than others. Simpler may for some purposes prove better in the long run.

]]>
Comment on [MADS] Parallel Coordinates by vlk http://hea-www.harvard.edu/AstroStat/slog/2009/mads-parallel-coordinate/comment-page-1/#comment-1020 vlk Tue, 11 Jan 2011 09:09:59 +0000 http://hea-www.harvard.edu/AstroStat/slog/?p=1786#comment-1020 You can use reCAPTCHA, see http://www.google.com/recaptcha and http://code.google.com/apis/recaptcha/docs/wordpress.html You can use reCAPTCHA, see http://www.google.com/recaptcha and http://code.google.com/apis/recaptcha/docs/wordpress.html

]]>
Comment on [MADS] Parallel Coordinates by upvc windows http://hea-www.harvard.edu/AstroStat/slog/2009/mads-parallel-coordinate/comment-page-1/#comment-1019 upvc windows Sun, 26 Dec 2010 23:43:26 +0000 http://hea-www.harvard.edu/AstroStat/slog/?p=1786#comment-1019 Hey! I know this is kinda off topic but I was wondering if you knew where I could get a captcha plugin for my comment form? I'm using the same blog platform as yours and I'm having difficulty finding one? Thanks a lot! Hey! I know this is kinda off topic but I was wondering if you knew where I could get a captcha plugin for my comment form? I’m using the same blog platform as yours and I’m having difficulty finding one? Thanks a lot!

]]>
Comment on Yes, please by brianISU http://hea-www.harvard.edu/AstroStat/slog/2010/yes-please/comment-page-1/#comment-1017 brianISU Wed, 22 Dec 2010 14:24:31 +0000 http://hea-www.harvard.edu/AstroStat/slog/?p=4270#comment-1017 I agree with this statement. If I were to have an extremely wide confidence interval, I would not be very confident with it. I agree with this statement. If I were to have an extremely wide confidence interval, I would not be very confident with it.

]]>
Comment on About by chasc http://hea-www.harvard.edu/AstroStat/slog/about/comment-page-1/#comment-1016 chasc Mon, 13 Dec 2010 22:53:37 +0000 http://hea-www.harvard.edu/AstroStat/slog/?page_id=7#comment-1016 Sorry for the problems with the RSS feed. At the risk of sounding trite, but truly, we are encountering technical difficulties. Sorry for the problems with the RSS feed. At the risk of sounding trite, but truly, we are encountering technical difficulties.

]]>
Comment on About by Ali http://hea-www.harvard.edu/AstroStat/slog/about/comment-page-1/#comment-1015 Ali Mon, 13 Dec 2010 22:36:49 +0000 http://hea-www.harvard.edu/AstroStat/slog/?page_id=7#comment-1015 Hi - Any chance of fixing the RSS feed; I'd really like to subscribe in google reader. Thanks. Hi – Any chance of fixing the RSS feed; I’d really like to subscribe in google reader. Thanks.

]]>
Comment on About by Gus http://hea-www.harvard.edu/AstroStat/slog/about/comment-page-1/#comment-978 Gus Thu, 08 Jul 2010 18:59:39 +0000 http://hea-www.harvard.edu/AstroStat/slog/?page_id=7#comment-978 hi, the ".../feed/rss/" URL no longer seems to work. the content that this URL returns is broken/incomplete HTML. google reader seems to have a cache of its posts but that ended at the end of 2009. So sometime around the 1st of the year 2010, WP started producing garbage RSS. hi, the “…/feed/rss/” URL no longer seems to work. the content that this URL returns is broken/incomplete HTML.

google reader seems to have a cache of its posts but that ended at the end of 2009. So sometime around the 1st of the year 2010, WP started producing garbage RSS.

]]>
Comment on Everybody needs crampons by vlk http://hea-www.harvard.edu/AstroStat/slog/2010/sherpa-cheat-sheet/comment-page-1/#comment-970 vlk Sun, 13 Jun 2010 03:11:44 +0000 http://hea-www.harvard.edu/AstroStat/slog/2007/sherpa-cheat-sheet/#comment-970 Thanks for the link to the Sherpa scripts page, Doug. The best way to become familiar with a new language is to see lots of simple examples. Also thanks for the other comments. I will probably include those comments by moving this page off to gWave shortly -- it is much easier to update and make changes there than in the blog format. Thanks for the link to the Sherpa scripts page, Doug. The best way to become familiar with a new language is to see lots of simple examples.

Also thanks for the other comments. I will probably include those comments by moving this page off to gWave shortly — it is much easier to update and make changes there than in the blog format.

]]>
Comment on Everybody needs crampons by Doug Burke http://hea-www.harvard.edu/AstroStat/slog/2010/sherpa-cheat-sheet/comment-page-1/#comment-969 Doug Burke Sat, 12 Jun 2010 22:16:48 +0000 http://hea-www.harvard.edu/AstroStat/slog/2007/sherpa-cheat-sheet/#comment-969 Vinay, Have you looked at the <a href="http://cxc.harvard.edu/sherpa/scripts/index.py.html" rel="nofollow">Sherpa scripts</a> page. For instance, some of the plot options are discussed in the <a href="http://cxc.harvard.edu/sherpa/threads/setplot_manual/index.py.html" rel="nofollow">plotting thread</a>, although I couldn't see the error bars explicitly addressed. Try <code> p = get_data_plot_prefs() print(p) p["yerrorbars"] = False </code> To change the units of the y error bars, use the set_analysis() routine, for instance <code> set_analysis("wave", type="counts") </code> Also, for error estimation, we suggest you use <code> conf() </code> rather than projection nowadays, and note that the error analysis will be done in parallel (so expect all your cores to get used up). And finally, <code> ,ahelp set_analysis </code> can also be used to avoid quoting the first argument by using a comma before the command. This is an ipython-ism, and can be used for other commands, e.g. <code> ,load_pha foo.pha </code> Doug Vinay,

Have you looked at the Sherpa scripts page. For instance, some of the plot options are discussed in the plotting thread, although I couldn’t see the error bars explicitly addressed. Try


p = get_data_plot_prefs()
print(p)
p["yerrorbars"] = False

To change the units of the y error bars, use the set_analysis() routine, for instance


set_analysis("wave", type="counts")

Also, for error estimation, we suggest you use


conf()

rather than projection nowadays, and note that the error analysis will be done in parallel (so expect all your cores to get used up).

And finally,


,ahelp set_analysis

can also be used to avoid quoting the first argument by using a comma before the command. This is an ipython-ism, and can be used for other commands, e.g.


,load_pha foo.pha

Doug

]]>