elladd concern

Hello,

I compute addition on an elliptic curve using two different ways,

Sometimes I get the same output, sometimes not.

(13:21) gp > EE=ellinit([0,0,0,-1,0])

%449 = [0, 0, 0, -1, 0, 0, -2, 0, -1, 48, 0, 64, 1728, Vecsmall([1]), [Vecsmall([192, 1])], [0, 0, 0, 0, 0, 0, 0, 0]]

(13:21) gp > elladd(EE, ellztopoint(EE, I/3), ellztopoint(EE, 1/5))

%450 = [-2.505868825963744299759862918821279271135, 23.94289238927766278258847950281292062950]

(13:21) gp > ellztopoint(EE,I/3+1/5)

%451 = [-3.128363320745408068195829880608990254457 - 5.812438339617489547314821239328032779531*I, 17.04223696249744063278395168524217210843 + 0.9250366401407313569925930547405905790341*I]

(13:23) gp > elladd(EE, ellztopoint(EE, 1/5), ellztopoint(EE, 1/3))

%458 = [3.572821508137712718878965341811999639182, -6.483396971715026203877059953349878879318]

(13:24) gp > ellztopoint(EE,1/3+1/5)

%459 = [3.572821508137712718878965341811999639182, -6.483396971715026203877059953349878879318]

(13:24) gp > ellztopoint(EE,I/3+I/5)

%460 = [-3.572821508137712718878965341811999639182, -0.007445832599336367120361417691599058776603]

(13:29) gp > elladd(EE, ellztopoint(EE, I/5), ellztopoint(EE, I/3))

%461 = [34.03024146804892398019238227713760810747, 0.004740183757855107656643668149420818213378]

What am I missing here ? It seems elladd or ellztopoints does not really stand complex points ?

Best,

Pierre C.