I compute the integral of modular forms using the pari command in sagemath as follows.
def phi(s,t): mf = pari.mfinit([27, 2], 0) f = pari.mfeigenbasis(mf)[0] symb = pari.mfsymbol(mf,f); F=(pari.mfsymboleval(symb, [s,t])2pi*I) return F
but the results show that phi(oo,0)+phi(0,a)=-phi(oo,a) where a is the cubic root of units in the upper half plane. Should it be phi(oo,0)+phi(0,a)=phi(oo,a) ? what is the reason?