THOR (Tracklet-less Heliocentric Orbit Recovery) links astronomical detections across multiple epochs to identify observations belonging to the same moving Solar System object. The software searches detection catalogs by testing heliocentric orbit hypotheses and associating observations consistent with each candidate orbit. This approach enables recovery of asteroids and other moving objects without requiring intranight tracklets or a predefined observing cadence. THOR can therefore analyze survey data sets with irregular temporal sampling and link observations separated by days to months. The code supports large-scale processing of survey catalogs to aid discovery and orbit determination of asteroids and other minor bodies.

APS finds Frequentist confidence limits on high-dimensional parameter spaces by using Gaussian Process interpolation to identify regions of parameter space for which chisquared is less than or equal to some specified limit. The code is written in C++, is robust against multi-modal chisquared functions and converges comparably fast to Monte Carlo methods. Code is also provided to draw Bayesian credible limits using the outputs of APS, though this code does not converge as well. APS requires the linear algebra libraries LAPACK, BLAS, and ARPACK (ascl:1311.010) to run.