Thanks Aai, Your definition gives the same results for the tests at the end of the article as Eugene's original, so we have a working version now.
I imagine it would be preferable to use Eugene's original code as closely as possible. Does anyone have a copy of the original Vector article (Vector 11, 3 (January 1995)) they can check to see what was printed there? Alternatively could some of the differences be explained by changes in J since J8? > From: Aai > May be this is of some help. > > Playing with combinations of mxp it seems that only(?) applying p=: 0 3 > 1 2 will lead to the same test results. By a small modification of mxp > you get the same result: > > mxp=: [:C. 1 0;<:@(,~ >:@|.)@>:@+:@i...@-: > > mxp 4 > 0 3 1 2 > > I don't know however if it will work for 6, 8 etc... > > > > =@@i > > > Hallo Sherlock, Ric, je schreef op 27-03-09 10:21: > > I've been working on the Jacobi's Method article, but have run in to > a problem with the definition of the verb mxp. > > > > >From the Word document: > > > > mxp=: [: C. 0: ; <: , (,~ >:@|.)@>:@+:@i.@<:v > > mxp 6 > > |value error: v > > | mxp 6 > > mxp1=: [: C. 0: ; <: , (,~ >:@|.)@>:@+:@i.@<: > > mxp1 6 > > |index error: mxp1 > > | mxp1 6 > > > > > > I suspect that the v on the end is a typo, but even then the verb > gives an index error. > > > > >From the description of what it is supposed to do: > > ============================================== > > The subverb mxp (make index permutation) takes a positive even > integer as argument and yields a list which is a permutation of the > integers from 0 through one less than the argument. The permutation is > such that when applied repeatedly to a conforming list, none of the > successive pairs in the lists are equal. > > =============================================== > > > > ... I came up with the following but I don't think it's quite right. > Any ideas? > > > > mxp2=: [: C. 0 ; (,~ <:@|.)@:>:@:+:@i...@-: > > mxp 6 > > 1 3 0 5 2 4 > > > > --------------------------------------------------------------------- > - > > For information about J forums see > http://www.jsoftware.com/forums.htm > > > ---------------------------------------------------------------------- > For information about J forums see http://www.jsoftware.com/forums.htm ---------------------------------------------------------------------- For information about J forums see http://www.jsoftware.com/forums.htm
