31

Is there a software (open source preferred) where I would input something like "Ingoing: a fermion $(p1, s1)$ and a photon $(p2, s2)$. Output: A fermion $(k1, r1)$ and a photon $(k2, r2)$" and I would could then get each diagram (that is, each term in the Dyson series) up to whichever order I choose, also showing terms that evaluate to zero (just to see that they are there)?

I have tried to compute the above diagram to 2nd order and I'm just baffled with how many details there are to collect, this seems like the perfect job for a computer, but I can't find anything that would do such a symbolic calculation (hopefully showing the steps along the way) and not just the final result.

Qmechanic
  • 201,751
PPR
  • 2,004
  • 3
    Maybe "FORM" and in particular http://www.feynarts.de/formcalc/. From wiki: The software package FormCalc which is widely used in the physics community to calculate Feynman diagrams is built on top of FORM. – Hunter Feb 03 '14 at 01:38
  • 2
    This question appears to be off-topic because it is about software recommendations and not physical concepts – tpg2114 Feb 03 '14 at 02:30
  • 3
    @tpg2114 Why would this questions be off-topic, whereas this is not off-topic? – Hunter Feb 03 '14 at 02:59
  • @Hunter: I'm not that familiar with the FormCalc package. However form the arXiv paper that I cite, it looks like a quite useful software. Nice suggestion! – Zoltan Zimboras Feb 03 '14 at 03:35
  • @ZoltanZimboras yeah, I have never used it myself, but my lecturer has mentioned it a couple of times as a useful program to use after one feels comfortable evaluating Feynman diagrams by hand. – Hunter Feb 03 '14 at 03:37
  • @Hunter Yes, I agree! That's something one should point out about all these packages. – Zoltan Zimboras Feb 03 '14 at 03:43
  • @Hunter That original question is from 2010 and our policies have changed since then. The policy is really that recommendation questions are off-topic because A) they are opinion based and B) they are really just requesting a list, and that's exactly the kind of answer you have so far. But -- I'm only one vote of the 5 required to close and if people don't agree with me, it won't get closed. – tpg2114 Feb 03 '14 at 04:02
  • @tpg2114 oh ok, thanks for letting me know; I wasn't aware of that. – Hunter Feb 03 '14 at 04:07
  • 1
    @tpg2114 and others, we agreed in meta that questions about software exclusively used by and useful for physicists purpose are allowed. So there is no need to close this question, voted to leave open. Please do not overrule that agreement like this. – Dilaton Feb 03 '14 at 08:05
  • 2
    @Hunter both questions are allowed according to the meta agreement I cited, the close votes are therefore not justified. – Dilaton Feb 03 '14 at 08:10
  • 1
    Hopefully you know how to compute simple diagrams by hand already. These codes are not quite "black boxes." – Tom Dickens Feb 03 '14 at 09:33
  • @Dilaton Before getting all bent out of shape, recognize that I'm the only one who voted to close so it's pretty clear it's not going to be closed. This is why it takes 5. – tpg2114 Feb 03 '14 at 15:42
  • @TomDickens: I completely agree with you! We even talked about this a bit above. Probably I should point this out in an edit to the answer. I'll do it now. – Zoltan Zimboras Feb 03 '14 at 16:53

1 Answers1

17

There are, of course, a lot of codes floating around. Which of them you should choose, depends on what you want to calculate exactly. Here I mention four possibilities:

  1. CALHEP - this package takes you from a given Lagrangian through its Feynmann rules to the calculation of cross sections.

  2. xloops - this package calculates the 1-PI Feynman diagrams with one and two closed loops for a given process in the Standard model and related theories.
    Note added: as pointed out in the comments, the link does not work (as of July, 2021), will update this paragraph when a new link will be available.

  3. You should also take a look at MadGraph.

  4. And here is a nice paper that discusses how to generate and calculate one-loop Feynman diagrams with three software packages (each used at different steps). For example, it discusses the FormCalc package of FORM which was also mentioned by Hunter in his comment.

I hope some of these will help you with the particular calculation that you want to perform.

Edit. Let me add here also the comment of TomDickens, because this point should be stressed: Hopefully you know how to compute simple diagrams by hand already. These codes are not quite "black boxes".

Zoltan Zimboras
  • 1,302
  • 1
    I wonder if programs applying twistor or even amplituhedron methods to evaluate Feynman diagrams are already available too ...? – Dilaton Feb 03 '14 at 08:49
  • 1
    @Dilaton: Good question! I don't know of any. Triggered by your question, I even made a (superficial) search on the web and didn't find such a software. I would also be interested whether such programs already exist. – Zoltan Zimboras Feb 03 '14 at 16:50
  • 1
    @ZoltanZimboras, thanks for your answer, I will check all these. What I have done so far is work through chapters 4 and 5 in Peskin. However, I felt Peskin glosses over many of the technicalities with all steps of the way and this is precisely why I would love to have such software to compare and see that indeed everything I think I understand I really understand correctly. – PPR Feb 03 '14 at 21:19
  • @Psycho_pr: I think this will be a really good exercise! I'm also really curious what you think about the different software packages. If you have time, let me know which you found useful and which you didn't (either by writing a comment here, or sending a mail). – Zoltan Zimboras Feb 03 '14 at 21:32
  • 1
    Unfortunately I still didn't get to test the extensive software available but I made a small addition of myself which applies to calculating Feynman propagator functional derivatives in Yukawa theory up to arbitrary orders. I reckon it can be extended to QED with not much effort: https://github.com/yashkes/manyDerivatives – PPR Apr 15 '14 at 21:46
  • I just clicked on your link, nice description of your problem and software (with link to the original homework problem). This is great!! Thanks for sharing with us. – Zoltan Zimboras Apr 16 '14 at 08:16
  • 1
    MadGraph does only tree level (no loops but many outgoing particles), MadGraph5_aMC@NLO can do one loop – Andre Holzner Nov 06 '14 at 20:47
  • 1
    The link 2) seems dead now. – Urb Jul 19 '21 at 17:48
  • @Urb - thanks a lot for your comment. I checked it and you are right, the link does not work anymore. I also tried to search whether a new link exists or not, but did not find it. So for the moment I updated the answer with a note about the current situation based on your comment. – Zoltan Zimboras Jul 21 '21 at 05:32