MathOverflow is a question and answer site for professional mathematicians. Join them; it only takes a minute:

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

Let $f_0(T),f_1(T) \in \mathbb R [T]$ be polynomials. Let $F(T,X):=X^2+f_1(T)X+f_0(T)$. Then the set $M(t):=\{x \in \mathbb R \mid F(t,x) \le 0\}$ is bounded. Its volume $V(t)$ is $\sqrt{\max\{0,f_1(T)^2-4f_0(T)\}}$. In particular this is piecewise continously differentiable and monotonous, and the number of pieces can be bounded from above in terms of the total degree of $F$.

This motivates the following much more general questions:

For $i=1, \dots, r$, let $F_i \in \mathbb{R}[T,X_1, \dots, X_n]$ be polynomials. We assume that for all real $t > 0$, the sets $M(t):=\{x=(x_1, \dots, x_n) \in \mathbb R^n \mid \text{$F_i(t,x_1, \dots, x_n) \le 0$ for all $i=1, \dots, r$}\}$ are bounded.

I am interested in the function $V(t)$ defined as the volume of $M(t)$, for $t > 0$. In particular:

  • Is $V(t)$ piecewise continuously differentiable, with finitely many pieces?
  • Is $V(t)$ piecewise monotonous, with finitely many pieces?
  • If so: Is there a bound on the number of pieces that depends only on $n$, $r$, the total degree of $F_1, \dots, F_r$?

I would also be interested in conditions on the $F_i$ under which there are such uniform bounds on the number of pieces.

share|cite|improve this question
up vote 2 down vote accepted

A theorem by Jean-Philippe Rolin and Jean-Marie Lion asserts a similar property in the analytic category. See Intégration des fonctions sous-analytiques et volumes des sous-ensembles sous-analytiques, Annales de l'institut Fourier (1998) 48 (3), p.755-767.

I quote their abstract : "Let $f(x,y)$ be a positive subanalytic function defined on $\mathbf R^n×\mathbf R^m$. We prove that the integral $\int_{\mathbf R^m}f(x,y)\,dy$ is a log-analytic function of $x$. Let $Y_x$ be a subanalytic family of global subanalytic subsets of the euclidean space $\mathbf R^m$. We deduce from the previous result that the $k$-dimensional volume of $Y_x$ is a log-analytic function of $x$. A corollary is the log-analytic behaviour of the $k$-dimensional density of a $k$-dimensional subanalytic set at any point of its topological closure."

share|cite|improve this answer

Your Answer


By posting your answer, you agree to the privacy policy and terms of service.

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