Higgs Combination Applet

I have been showing unofficial Higgs combinations here for the last year or so but maybe you want to try some unusual combinations of your own. Now you can using the viXra unofficial Higgs combination Java applet. It is armed with most of the plots published by the experiments CDF, D0, CMS, ATLAS and LEP. You just have to choose how to combine them. I am hoping it is self-explanatory but ask some questions and you may get some good tips. You may need to update your Java plug-in.

Disclaimer: The results are approximate, unofficial and not endorsed by the experiments.

17 Responses to Higgs Combination Applet

  1. JollyJoker says:

    Very nice, thanks!

    Is there a simple explanation why Custom sub-channel with all selected gives such a different result from Unofficial? Are the subchannels LHC only?

    • Philip Gibbs says:

      There are a couple of things that are not listed under sub-channels, some bb and LEP. I will add them in later to make it more consistent. Use the “List” option under plot type to see exactly what is included under each selection and you will understand better. There could easily be bugs.

  2. Lubos Motl says:

    Great work, Phil! Kind of amazing that there’s no official applet that would dream about competing against you. 😉

  3. Lubos Motl says:

    Thanks to your applet, I learned that there’s Java 7 instead of Java 6, so I upgraded. 😉

    Looking at the options, why does “Lumi” offer 10.5 aside from 10? Isn’t it too close? Didn’t you mean to include 12.5 in the list of options? 😉

    • Philip Gibbs says:

      I did that at some point when the Tevatron finished with 10.5/fb. The purpose of the option is to scale up the contributions so that you can see what the size of the errors will be when they reach that amount of data (you probably already worked that out)

  4. anna v says:

    Why don’t you give access to it from your toolbar, after “science calendar” ? That way when this post falls off the first page it will be easily accessible without search.

    • Philip Gibbs says:

      Thise have to be wordpress pages I think, but I have added it as a link on the right

  5. Is it released as an open source software? Great app anyway, you should port it on android/iOS asap!

    • Philip Gibbs says:

      What and let other people muck about with my lovely code? 🙂 Don’t android and iOS have web browsers then? 🙂 Seriously though, thanks for the complement.

  6. Yasmin says:

    Great and helpful app, Phil ! 🙂 Will be very helpful for my next scientific talk on the Higgs in June, esp. for non-accelerator colleagues. Your app really shows how many parameters are used for getting a result, but in most of those I already tried, the trend clearly shows “something” instead of “nothing”. What about different backgound options or are they already included and I did not find them ?

    • Philip Gibbs says:

      Thanks, glad it is of help. Do you mean you want to be able to vary/remove the background sigma bands, or are you asking about the way experimental background is included?

      • Yasmin says:

        In connection with the general comparison of (i) the December 2011 (4TeV) results with those of 2012 (7TeV) and (ii) those of 2012 (7TeV) from day to day in your app, I thought not only about (a) the machine induced background depending on the machine operating conditions / beam conditions but also (b) the rate of the background from the beam–gas losses dep. directly on the residual gas pressure in the detector sections, and (c) changes of parameters of the machine operation affecting the total background levels and background formation and/or (d) other effects I did not mention here.
        Esp. with regard to (i), I know that lower beam lifetime will result
        in a significant increase of background. Thus, comparison of the (i) results shall be difficult and/or complex.
        But what about (existing) background variations of the (ii) results ? And, finally, what may this mean for the app, i.e., the self-analysis of the results ? Or, do you think this is of minor importance for the use of the app (and the input of a generalized background is sufficient) ?
        Hope you understand my point – this is just for explaining to myself, if and what impact this may have.
        Thanks for any further comment.

      • Philip Gibbs says:

        There are many ways in which the running conditions change all the time. Energy, luminosity, trigger algorithms etc. Only experts within the collaborations know enough about these details to analyse them. They have to simulate the effect of each change using monte carlo methods and compare that with observation.

        However, the exclusion plots that they end up with give an answer that just tells us the probability of any observed excess according to whether or not a SM Higgs boson exists with a given mass. If we assume that uncorrelated normal distributions are a good approximation then just simple arithmetic based on the laws of probability are required to do the combination. The assumption is not pefect but comparison with official calculations has shown it to be good enough to be worth doing.

        In other words, all the things you mention have been factored in by the analysts from the experiments and the applet does not need to know about them to a first order approximation.

  7. JollyJoker says:

    After ICHEP we’re going to be less interested in comparing to no-Higgs background and more in comparing to background plus an SM Higgs. For the full combination that should work just using your gaussian fit, but can you use the gaussian Higgs fit of the full combination in single channel plots?

    • Philip Gibbs says:

      Where there are enough events the unofficial channel combinations will work. The last time they gave official combinations was at HCP11 with up to 2/fb and if you compare the official/unofficial combos for diphoton at that time you will find that the result is pretty good. The ZZ->4l channel does not come out so well but the signal plot is better than the exclusions plot and that is the one you need. With 5/fb or more I think it will be fine for the main channels.

      • JollyJoker says:

        How about the width of the gaussian? If you plot what is expected for an SM Higgs @ 125 GeV and take the difference between that and the observed curve, how do you know the uncertainty in mass for different channels? Or is this a stupid question and that’s just allowed to vary with the data?

      • Philip Gibbs says:

        Oh you mean the yellow fitted gausian on the signal plot. The width of the signal is controlled by the detector resolution so it is not of any physical interest. The real width of the resonance is much narower and even the ILC wont be able to resolve it. All we can look at are the branching ratios and (later) production cross-sections

%d bloggers like this: