Comparisons of exact numbers

*To*: mathgroup at yoda.physics.unc.edu*Subject*: Comparisons of exact numbers*From*: JWENDEL at isdmnl.wr.usgs.gov*Date*: Wed, 19 Aug 1992 20:22:49 -0700 (PDT)

As discussed by Roger B. Kirchner, <kirchner at cs.umn.edu> and David Jacobson, <jacobson at cello.hpl.hp.com>, perhaps it should be added that using N[...] may not be the best strategy. In Kirchner's example, find Min[2^(1/2),3^(1/2)], it would surely be better to square both terms and compare the results. Here's another: find Min[x,y] where x=Sqrt[5]+Sqrt[22+2Sqrt[5]]; y=Sqrt[11+2Sqrt[29]]+Sqrt[16-2Sqrt[29]+ 2Sqrt[55-10Sqrt[29]]]; .........Jim Wendel