A speed-up of the algorithm for computing comprehensive Gröbner systems

Abstract

We introduce a new algorithm for computing comprehensive Gröbner systems.There exists the Suzuki-Sato algorithm for computing comprehensive Gröbner systems. The Suzuki-Sato algorithm often creates overmuch cells of the parameter space for comprehensive Gröbner systems. Therefore the computation becomes heavy. However, by using inequations ("not equal zero"), we can obtain different cells. In many cases, this number of cells of parameter space is smaller than that of Suzuki-Sato's. Therefore, our new algorithm is more efficient than Suzuki-Sato's one, and outputs a nice comprehensive Gröbner system. Our new algorithm has been implemented in the computer algebra system Risa/Asir We compare the runtime of our implementation with the Suzuki-Sato algorithm and find our algorithm superior in many cases.