So I have installed all FeynArts folders (from 3.5 to 3.7) with FeynCalc 8.0.1., and none of it is working properly :\

First I load the path to specific FeynArts package (3.5 for instance), then, by:

<< HighEnergyPhysics`FeynCalc`;

I load FeynCalc, and he says

Loading FeynCalc from C:\Users\*****\AppData\Roaming\Mathematica\Applications\HighEnergyPhysics
FeynCalc 8.1.0 For help, type ?FeynCalc, open FeynCalcRef8.nb or visit www.feyncalc.org
Loading FeynArts, see www.feynarts.de for documentation
FeynArts 3.5 patched for use with FeynCalc

So everything seems nice, right? Wrong!

I even load FeynArts package, and no errors. But when I want to initialize the proper model I get:

In[9]:= InitializeModel[SM]

During evaluation of In[9]:= 

During evaluation of In[9]:= loading generic model file C:\Users\*****\AppData\Roaming\Mathematica\Applications\HighEnergyPhysics\FeynArts-3.5\Models\Lorentz.gen

During evaluation of In[9]:= TagSetDelayed::tagnf: Tag FourVector not found in \!
\(TraditionalForm\`\(-\(TraditionalForm\`mom_\)\^\(TraditionalForm\`\(TraditionalForm\`\(Pattern[\(\(mu___, _\)\)]\)\)\)\)\)
. >>

During evaluation of In[9]:= Pattern::patvar: First element in pattern Pattern[b_,_] is not a valid pattern name. >>

During evaluation of In[9]:= Pattern::patvar: First element in pattern Pattern[b_,_] is not a valid pattern name. >>

Out[9]= $Aborted

And I can't do anything! I can plot topology, but I can't insert fields. I've been doing everything by the book :\ What seems to be wrong? :\

  • $\begingroup$ Could you described exactly what you did during installation? Your description "installed all FeynArts folders (from 3.5 to 3.7) with FeynCalc 8.0.1" sounds weird to me. Anyway, Rolf Mertig who has been working on this package is a regular visitor here, so I assume he'll drop in once he sees this question. $\endgroup$ Commented Jul 6, 2012 at 9:55
  • $\begingroup$ Well I downloaded the files, FeynArts from here: feynarts.de, and FeynCalc from here: feyncalc.org/download/index.html, and I put them in my C:\Users*****\AppData\Roaming\Mathematica\Applications folder. FeynArts inside FeynCalc folder... I have a feeling that FeynCalc is in collision with FeynArts package :\ $\endgroup$
    – dingo_d
    Commented Jul 6, 2012 at 10:00
  • 5
    $\begingroup$ Actually I did not work on FeynArts, but FeynCalc. Frederik Orellana, the other mainainer of FeynCalc is working on a better integration of a newer FeynArts. I will send him an email right now.And yes, there is some type of trouble if you install the packages independently, this is why FeynArts has to be patched before it can be used together with FeynCalc. However, if you are in a hurry you can use each package in a separate Mathematica run. $\endgroup$ Commented Jul 6, 2012 at 10:53
  • 3
    $\begingroup$ @dingo_d Of course, you always able to associate different MathKernel process with different Notebooks in the same Mathematica run (see the Evaluation :: Notebook's kernel menu). As I remember, it is even possible to set separate kernels for separate Cells in one Notebook through the Evaluator option of Cell. $\endgroup$ Commented Jul 6, 2012 at 14:11
  • 1
    $\begingroup$ If you use this package a lot, you can consider posting about it here, for more exposure: meta.mathematica.stackexchange.com/questions/428/… $\endgroup$
    – Szabolcs
    Commented Jul 9, 2012 at 10:46

2 Answers 2


I've just uploaded a new version which again should support loading FeynArts, so that FeynCalc and FeynArts can be used from the same session - which I personally find very convenient.





FeynArts-3.7 is bundled with the release, so you can simply load FeynCalc with

$LoadPhi = True;
    $LoadFeynArts = True;
<< HighEnergyPhysics`FeynCalc`;

The first time you do this, FeynArts will be patched to avoid name clashes etc. After that you can in principle load FeynCalc+FeynArts without loading PHI, but then you should manually do

SetOptions[FourVector, FeynCalcInternal -> False];



Otherwise FeynArts will complain and not load correctly

because of

FourVector /: -FourVector[mom_, mu___] := FourVector[Expand[-mom], mu]

in "Lorentz.gen".

  • 2
    $\begingroup$ Thanks! And welcome to Mathematica.SE! $\endgroup$ Commented Jul 7, 2012 at 15:52
  • $\begingroup$ Thank you very much ^^ This will help immensely :) $\endgroup$
    – dingo_d
    Commented Jul 7, 2012 at 21:48
  • $\begingroup$ Oh, and I've tried it, and it works perfectly :) I am very grateful :) $\endgroup$
    – dingo_d
    Commented Jul 8, 2012 at 6:38
  • $\begingroup$ I also want to say Thank you! I am currently trying to start using FeynCalc for some Amplitudes... and well, wow, it is really difficult to get started. I don't even know what I will do if I have to use our Lagrangian, but well, I'll figure it out :) But if I also had had to fix that problem above, it would take some more days for me to get finally started. So thank you all, for providing us with such great support!!! $\endgroup$
    – user4632
    Commented Nov 13, 2012 at 11:45
  • $\begingroup$ @MlleHaberlienzer You can find some simple lagrangians in Lagrangian.m (feyncalc.org/FeynCalcBook/Lagrangian) and with FeynRule (feyncalc.org/FeynCalcBook/FeynRule) you can derive Feynman rules. $\endgroup$ Commented Nov 13, 2012 at 12:15

Just in case someone stumbles upon this old question. Since FeynCalc 9.x, patching of FeynArts is done automatically when you install FeynCalc. Just evaluate


and let the automatic installer do the job.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.