From: Robert D. <rob...@us...> - 2005-01-10 02:50:42
|
Update of /cvsroot/maxima/maxima/archive/share/trash In directory sc8-pr-cvs1.sourceforge.net:/tmp/cvs-serv24472/archive/share/trash Added Files: airy.usg Log Message: Move share/specfunctions/airy.usg to archive/share/trash/airy.usg. (Useful content of airy.usg has been merged into doc/info/Special.texi.) --- NEW FILE: airy.usg --- May 2, 1981 4:52 pm by Leo P. Harten (LPH) The Airy equation diff(y(x),x,2)-x*y(x)=0 has two linearly independent solutions, taken to be Ai(x) and Bi(x). This equation is very popular as an approximation to more complicated problems in many mathematical physics settings. Do LOAD("AIRY"); to get the functions AI(x), BI(x), DAI(x), DBI(x) . The file SHARE1;AIRY FASL (by LPH@MIT-MC) contains routines to compute the Ai(x), Bi(x), d(Ai(x))/dx, and d(Bi(x))/dx functions. The result will be a floating point number if the argument is a number, and will return a simplified form otherwise. An error will occur if the argument is large enough to cause an overflow in the exponentials, or a loss of accuracy in sin or cos. This makes the range of validity about -2800 to 1.e38 for AI and DAI, and -2800 to 25 for BI and DBI. The GRADEF rules are now known to MACSYMA: diff(AI(x),x)=DAI(x), diff(DAI(x),x)=x*AI(x), diff(BI(x),x)=DBI(x), diff(DBI(x),x)=x*BI(x). The method is to use the convergent Taylor series for abs(x)<3., and to use the asymptotic expansions for x<-3. or x>3. as needed. This results in only very minor numerical discrepancies at x=3. or x=-3. More accuracy can be had if you request from LPH. For details, please see Abramowitz and Stegun's Handbook of Mathematical Functions, section 10.4 (hardcover ed.) and Table 10.11 . To get the floating point Taylor expansions of the functions here, do ev(TAYLOR(AI(x),x,0,9),infeval); for example. Please also try SHARE;BESSEL FASL (by CFFK) for the AIRY function there. Leo P. Harten (LPH) |