search

JP-Ellis / tikz-feynman

Feynman Diagrams with TikZ
http://www.jpellis.me/projects/tikz-feynman
Other
146 stars 20 forks source link

Wrong sign in xshift of node in `with reversed arrow`? #48

Open TorbjornT opened 6 years ago

TorbjornT commented 6 years ago

This is in reference to https://tex.stackexchange.com/questions/377299/tikz-feynman-asymmetry-between-particle-and-antiparticle-lines/410367#410367

Should the xshift of the node in with reversed arrow have the opposite sign as that in with arrow? Currently both with arrow:

https://github.com/JP-Ellis/tikz-feynman/blob/8d1859030b705d8f0fb9750ec737ea47cfa0f30f/tikzfeynman.keys.code.tex#L295

and with reversed arrow:

https://github.com/JP-Ellis/tikz-feynman/blob/8d1859030b705d8f0fb9750ec737ea47cfa0f30f/tikzfeynman.keys.code.tex#L313

do xshift=-0.5mm, causing the asymmetry mentioned in that TeX.SX post.

losmaxos commented 6 years ago

Hi, seems like this is due to TikZ not placing the isoscele triangles "half-width-center" at the lines' center, but rather at their center of gravity or something like this. This roughly needs an xshift of 1/6 the arrows width, see below. In addition to the symmetry desired by the original poster (arrows point in SAME direction), we also want symmetry when we have a fermion and an antifermion right above each other, with the arrows pointing in OPPOSITE directions.

What worked for me was using for with arrow:

      xshift= -0.166666*arrow width,

and for reversed arrow:

      xshift= 0.166666*arrow width,