A class of complete potential-density basis sets in cylindrical (R,phi,z) coordinates is presented. This class is suitable for stability studies of galactic disks in three dimensions and includes basis sets tailored for disks with vertical density profiles that are exponential (exp(-|z|/\zn)), Gaussian (exp(-(z/\zn)^2) or locally isothermal (sech^2(z/\zn)). The basis sets are non-discrete and non-biorthogonal; however, the extra numerical computations required (compared with discrete biorthogonal sets) are explained and constitute a small overhead. The method of construction (and proof of completeness) is simple and can be used to construct basis sets for other density distributions that are best described in circular or elliptic cylindrical coordinates. When combined with a basis set designed for spheroidal systems, the basis sets presented here can be used to study the stability of realistic disks embedded in massive halos.

The stability of a galaxy model is most easily assessed through N-body simulation. Particle-mesh codes have been widely used for this purpose, since they enable the largest numbers of particles to be employed. We show that the functional expansion technique, originally proposed by Clutton-Brock for other simulation problems, is in fact superior for stability work. For simulations of linear evolution it is not much slower than grid methods using the same number of particles, and reproduces analytical results with much greater accuracy. This success rests on its ability to represent global modes with a modest number of basis functions; grid methods may be more effective for other applications, however. Our conclusions are based on implementations of functional expansion and grid algorithms for disk galaxies.

Most numerical integration algorithms are not designed specifically for Hamiltonian systems and do not respect their characteristic properties, which include the preservation of phase space volume with time. This can lead to spurious damping or excitation. Methods that do preserve all the Hamiltonian properties, i.e., for which the time-forward map is symplectic, are called symplectic integration algorithms or SIAs. Although such integrators are symplectic in theory, they are not symplectic if implemented using finite-precision arithmetic. This paper explains how to eliminate this problem by using ``Lattice SIAs'' and shows that these methods yield significant advantages when the computational error is dominated by roundoff. Using a lattice SIA is equivalent to evolving the exact solution of a problem with a Hamiltonian that is slightly different from the original. Lattice methods are useful for studies of the long-term evolution of Hamiltonian dynamical systems.