90 3. Factors and Zeros(15) Show thatUT + up + *.. + 26; _ (?/*+I+ 1+ y-(*+‘))(y” + 1+ y-“) - 3(t + 1)
u1+ ‘u2 + *. * + U” - (t2 - 4)@ + 1)Deduce from this that UT + uz +... + uz is divisible by u1 + u2 +.. .u,
(1 5 n).
Define polynomials v, and w, as follows:VI-J = 0 Vl = 1V2m = VZm-1 + Vzm-2 (m^2 1)
V2mtI = (t - 2)vzm + V2m-1 (m 1 1)wo = 2 Wl = 1
w2, = (t - 2)2~2~-1+ wzm-2 (m 1 1)
W2m-t1 = w2m + warn-1 (m 2 1)(16) v,+z = tv, - vn-2 wn+z = ha - w-2 (n L 2)(17) v2m = urn (m 2 0)(18) vzm+l(t) = (-l)*wmtl(-~) (m 2 0)(19) U” = vnwn = ~(v”+lw”-1+ ~“-1wntl)(20) When n = 2’, u, = plp2. “p, where pi(t) = t and pk = pzBl -^2
(k 2 1).3.3 Finding Integer and Rational Roots: Newton’s
Method of DivisorsA first step in factoring polynomials over Q is to use the Factor Theorem
to locate linear factors by finding rational zeros. It is straightforward to
see that all but finitely many rationals can be rejected as possible zeros.
Special techniques will narrow down the possibilities even further.
Exercises
In these exercises, q(t) will represent the polynomial
cntn + C,-ltn-l + C,-p--2 $ *.. + qt + cowith integer coefficients ci.