Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP Posting-Version: version B 2.10.3 4.3bsd-beta 6/6/85; site batcomputer.TN.CORNELL.EDU Path: utzoo!watmath!clyde!burl!ulysses!bellcore!decvax!ittatc!dcdwest!sdcsvax!sdcrdcf!hplabs!tektronix!uw-beaver!bullwinkle!batcomputer!goon From: goon@batcomputer.TN.CORNELL.EDU (Jon Goon) Newsgroups: net.math.symbolic Subject: macsyma pattern-matcher bug? Message-ID: <388@batcomputer.TN.CORNELL.EDU> Date: Tue, 11-Mar-86 11:02:06 EST Article-I.D.: batcompu.388 Posted: Tue Mar 11 11:02:06 1986 Date-Received: Sat, 15-Mar-86 22:19:15 EST Reply-To: goon@batcomputer.UUCP (Jon Goon) Organization: Theory Center, Cornell University, Ithaca NY Lines: 47 Howdy. I am having some trouble using the pattern-matching facility in macsyma. Any hints on what I am doing wrong? The sequence matchdeclare(nn,true); defrule(pat,cos(nn*t),-sin(nn*t)/nn); applyb1(cos(a*t),pat); gives the correct answer, -sin(a*t)/a. However, if I try matchdeclare(nn,true); defrule(pat1,%e^(nn*t),%e^(nn*t)/nn); then macsyma echoes the correct rule, nn t nn t %e pat1 : %e -> ------- nn but when I apply it with applyb1(%e^(t),pat1); I get garbage: t t t %e. I get the same answer if I express everything in terms of the exp function. Any reason that this should happen? Incidentally, I am running macsyma 304 on a dec 2065. |----------------------------------------------------------------------------| | J. Goon | | Center for Applied Math. | | Cornell U. | | Ithaca, N.Y. 14853 | | {decvax,ihnp4,cmcl2,vax135}!cornell!amvax!goon (USENET) | | goon%amvax@CRNLCS.BITNET (Bitnet) | | goon@amvax.tn.cornell.edu (ARPANET) | |----------------------------------------------------------------------------|