In this paper we talk about a new efficient numerical approach to deal with
inaccuracy when implementing geometric algorithms. Using various floating-
point filters together with arbitrary precision packages, we develop an eas
y-to-use expression compiler called EXPCOMP. EXPCOMP supports all common op
erations +, -, ., /, root. Applying a new semi-static filter, EXPCOMP combi
nes the speed of static filters with the power of dynamic filters. The filt
er stages deal with all kinds of floating-point exceptions, including under
flow. The resulting programs show a very good runtime behaviour.