This could be primary school stuff. But I want to ask it.

In factoring $x^2+bx+c$ (i.e. $a = 1$ in $ax^2+bx+c$), we find $m$ and $n$ such that $m+n = b$ and $mn=c$. We can reason this well as follows:

In the expansion of $(x+m)(x+n)$, $mx$ and $nx$ should add up to $bx$ and $mn$ should result in $c$.

However I am not able to come up with the logic for following: For factoring $qx^2+bx+c$ (i.e. $a \neq 1$ in $ax^2+bx+c$), we find $m$ and $n$ such that $m+n=b$ and $mn = ac$.

  • I am not able to realize the logic behind why it is $mn = ac$ and
  • how doing this result in correct factors

The AC-method reduces to factoring a polynomial that is $\,\rm\color{#c00}{monic}\,$ (lead coeff $\color{#c00}{=1})$ as follows

$$\begin{eqnarray} \rm\: a\:f(x)\:\! \,=\,\:\! a\:(a\:x^2 + b\:x + c) &\,=\,&\!\!\rm\: \color{#c00}{X^2} + b\:X + \!\!\!\!\!\!\!\!\!\!\!\!\!\!\!\!\!\!{\overbrace{ac}^{\rm\qquad\ \ \ \ \ {\bf AC-method}}}\!\!\!\!\!\!\!\!\!\!\!\!\!\!\!\! =\, g(X),\ \ \ X = a\:x \\ \end{eqnarray}$$

For example

$$ {\begin{eqnarray} f \, &\,=\,& \ \, 2\ x^2-\ 3\ x\,\ +\ \ 1\\ \Rightarrow\ 2f\, &\,=\,&\!\ (2x)^2\! -3(2x)+2\\ &\,=\,& \ \ \ \color{#c00}{X^2}-\, 3\ X\,\ +\,\ 2,\,\ \ X\, =\, 2x\\ &\,=\,& \ \ (X-2)\ (X-\,1)\\ &\,=\,& \ (2x-2)\,(2x-1)\\ \Rightarrow\ \ f\:=\: 2^{-1}\,(2f)\, &\,=\,& \ \ \, (x- 1)\ (2x\,-1)\\ \end{eqnarray}}$$

If we denote our factoring algorithm by $\,\cal F,\,$ then the above transformation is simply

$$\cal F f\, = a^{-1}\cal F\, a\,f\quad\,$$

Thus we've transformed by $ $ conjugation $\,\ \cal F = a^{-1} \cal F\, a\ \,$ the problem of factoring non-monic polynomials into the simpler problem of factoring monic polynomials. This is sometimes called the AC method (cf. below). It works for higher degree polynomials too: we can reduce the problem of factoring a non-monic polynomial to that of factoring a monic polynomial by scaling by a $ $ power of the lead coefficient $\rm\:a\:$ then changing variables: $\rm\ X = a\:x$

$$\begin{eqnarray} \rm\: a\:f(x)\:\! \,=\,\:\! a\:(a\:x^2 + b\:x + c) &\,=\,&\!\!\rm\: X^2 + b\:X + \!\!\!\!\!\!\!\!\!\!\!\!\!\!\!\!\!\!\smash[t]{\overbrace{ac}^{\rm\qquad\ \ \ \ \ {\bf AC-method}}}\!\!\!\!\!\!\!\!\!\!\!\!\!\!\!\! =\, g(X),\ \ \ X = a\:x \\ \\ \rm\: a^{n-1}(a\:x^n\! + b\:x^{n-1}\!+\cdots+d\:x + c) &\,=\,&\!\!\rm\: X^n\! + b\:X^{n-1}\!+\cdots+a^{n-2}d\:X + a^{n-1}c \end{eqnarray}$$ After factoring the monic $\rm\,g(X)\, =\, a^{n-1}f(x),\,$ we are guaranteed that the transformation reverses to yield a factorization of $\rm\:f,\ $ since $\rm\ a^{n-1}$ must divide into the factors of $\rm\ g\ $ by Gauss' Lemma, i.e. primes $\,p\in\rm\mathbb Z\,$ remain prime in $\rm\,\mathbb Z[X],\,$ so $\rm\ p\ |\ g_1(x)\:g_2(x)\,$ $\Rightarrow$ $\,\rm\:p\:|\:g_1(x)\:$ or $\rm\:p\:|\:g_2(x).$

This method also works for multivariate polynomial factorization, e.g. it applies to this question.

See this answer for more on the ring-theoretic concepts that lie at the heart of the matter (primal elements, Schreier refinement, Riesz interpolation, etc).


If $m,n $ are the roots of $ax^2 + bx +c$ then we may write

$$ax^2 + bx +c = \color{red} d\ (x- m)(x-n) = dx^2 -d(m+ n)x + dmn$$

Using the equality of polynomials we have taht

$$d = a ; d(m+n) = -b ; mnd = c $$

that is

$$a = d ; m+n = \color {red}{-\frac{b}{a}} ; mn = \color{red}{\frac{c}{a}}$$

  • $\begingroup$ I was taught a different method, where you seek $m$ and $n$ that add to $b$ and multiply to $ac$. (However, the next step is not to write it in the form $a(x+m)(x+n)$, since your post shows that this doesn't work.) $\endgroup$ – Akiva Weinberger Mar 18 '15 at 23:23
  • $\begingroup$ If you have $2x^2+17x+21$, you want them to sum to 17 and multiply to 42; 14 and 3 work nicely. So you split it into $(2x^2+14x)+(3x+21)$. Factoring gives you $2x(x+7)+3(x+7)$, and factoring again gives you $(2x+3)(x+7)$. $\endgroup$ – Akiva Weinberger Mar 18 '15 at 23:27
  • $\begingroup$ It seems a litte tedious. What I presented is known as the Newton-Girard formulae. Both ways are, of course, the same. $\endgroup$ – Aaron Maroja Mar 18 '15 at 23:29
  • $\begingroup$ But then you have to deal with fractions. (By the way, rewriting it as $(2x^2+3x)+(14x+21)$ also works, if you were wondering.) $\endgroup$ – Akiva Weinberger Mar 18 '15 at 23:31
  • $\begingroup$ Yes, this may be a downpoint. But you see this again in generalized in Galois Theory. So I thought it was worth mentioning. $\endgroup$ – Aaron Maroja Mar 18 '15 at 23:33

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.