Yang, Ke RVarga, ZoltanParker, Kelsey AShu, YinanTruhlar, Donald G2022-02-142022-02-142022-02-14https://hdl.handle.net/11299/226289The PIPFit 2022 program can be used to develop analytic representations of potential energy surfaces for three-body and four-body systems. A weighted least-squares fit is performed with permutationally invariant polynomials (PIPs) whose variables are Morse-like bond functions, Gaussians, mixed exponential–Gaussians (MEGs), or hyperbolic secant variables. Three kinds of fit can be performed with the program:*PIPs fit to the whole potential, as originally proposed by Braams, Bowman, and Xie,*connected PIPs fit to the whole potential after removing the unconnected terms,*connected PIPs fit to the many-body part of the potential after removing the unconnected terms and the two-body terms.The program can also perform a two-stage fit in which one first fits lower-level energetic data with a large number of geometries and then fits higher-level corrections with a smaller set of geometries.License for software: Apache License, Version 2.0License for manual: CC-BY-4.0https://www.apache.org/licenses/LICENSE-2.0https://creativecommons.org/licenses/by/4.0/potential energy surface fittingpermutationally invariant polynomialsthree-body atomic systemsfour-body atomic systemsDatasethttps://doi.org/10.13020/tnzc-nt34