Name: Vladyslav Shtabovenko Date: 06/10/17-09:33:09 PM Z


Hi,

sorry for the late reply. What about the following:

exp2 = Collect2[FCI[exp], Momentum, IsolateNames -> KK] // Simplify

res = FVSeries[exp2, {q, 0, 4}]; // AbsoluteTiming

where exp is your original expression.

This requires 3.7 seconds on my machine with FeynCalc 9.2

Doing the expansion directly via

res = FVSeries[exp, {q, 0, 4}]; // AbsoluteTiming

takes 19.4 seconds, i.e. almost 5 times longer.

If 3.7 seconds are still too slow for you, I guess you would need
to go with FORM.

Cheers,
Vladyslav

Am 07.06.2017 um 13:04 schrieb Philipp:
> Hello,
>
> I have actually done a similar thing just that I have computed the FourDivergences one after another to not derive already existing expressions again. But that is probably as fast as it gets right? Because I have expressions much longer than the said one like
>
*> -(CA*(2*FVD[p1, Lor1] - FVD[q, Lor1])*(2*FVD[p1, Lor2] - FVD[q, Lor2])*SD[a, b]*SMP[“g_s”]^6*(FAD[{p1, I*mE}, p2, {p1 - q, I*mE}, {p1 - q, I*mE}]*((-(D*(psi[4] + 2*psi[13] - 2*psi[15])) - psi[26] - 2*psi[34] + CA^2*(omega[26] + D*(omega[4] - 2*(psi[4] + psi[13] - psi[15])) - 2*(psi[26] + psi[34] - psi[35])) - 2*CA*CF*(D*(psi[4] + 2*psi[13] - 2*psi[15]) + psi[26] + 2*psi[34] - 2*psi[35]) + 2*psi[35])*SPD[p1, p1] + 2*(D*(psi[4] + 2*psi[13] - 2*psi[15]) + psi[26] + 2*psi[34] - CA^2*(omega[26] + D*(omega[4] - 2*(psi[4] + psi[13] - psi[15])) - 2*(psi[26] + psi[34] - psi[35])) + 2*CA*CF*(D*(psi[4] + 2*psi[13]