1: \begin{abstract}

2: We consider a generalisation of Ulam's method for approximating

3: invariant densities of one-dimensional chaotic maps. Rather than use

4: piecewise constant polynomials to approximate the density, we use 

5: polynomials of degree $n$ which are defined by the requirement that 

6: they preserve the measure on $n+1$ neighbouring subintervals. Over the

7: whole interval, this results in a discontinuous piecewise polynomial

8: approximation to the density. We prove error results where this approach

9: is used to approximate smooth densities. We also consider the computation

10: of the Lyapunov exponent using the polynomial density and show that the

11: order of convergence is one order better than for the density itself.

12: Together with using cubic polynomials in the density approximation, this

13: yields a very efficient method for computing highly accurate estimates

14: of the Lyapunov exponent.  We illustrate the theoretical findings

15: with some examples.

16: \end{abstract}

17: