Localization of point sources represents an integral component of microscopy data analysis in applications such as the high-accuracy tracking of single molecules and the high-resolution visualization of subcellular structures labeled with stochastically activated fluorophores. The choice of a suitable method for localization and the customization of the chosen method are both critically dependent on various factors. These factors include the characteristics of the data such as the level of the signal detected from the point source and the types and amounts of noise that are present, experimental design choices such as the optics of the microscope used and the emission wavelength of the fluorophore used, and analysis requirements such as the desired processing throughput and level of localization accuracy. Consequently, the determination and customization of a localization method necessitate experimentation with various considerations, including the underlying optimization algorithm to use, the point spread function model with which to fit a point source, and the computational and algorithmic settings that affect the performance of the localization. As there are numerous combinations to evaluate, software is needed that enables one to efficiently carry out this experimentation. We describe here a software framework and implementation that addresses this important aspect of localization analysis. As a demonstration of this software, we use it to explore ways to improve the throughput of single molecule localization without sacrificing the localization accuracy. In doing so, we also highlight tools in the software that importantly allow the examination of localization results in great detail.