While writing a tacit version for this continued fraction task I noticed the ending comment:
"convince yourself that, in the same way as 3.7 may be represented as 3.70when an extra decimal place is required, [3;7] may be represented as [image: [3;7,\infty]] when an extra term is required" J knows that: 3 7 =&((+ %)/) 3 7 _ 1 I am careful when dealing with _ but in this case (as in many other cases) J's handling of _ makes a lot of sense; I did not see anything similar in any of the other implementations. By the way, for _151r77 J and Tcl reported a different representation versus the other languages reported representation; apparently both representations are valid: _2 25 1 2x ,&((+ %)/) _1 _1 _24 _1 _2x _151r77 _151r77 On Thu, Apr 18, 2013 at 11:19 PM, David Ward Lambert <[email protected] > wrote: > Please modify these new entries as you please: > http://rosettacode.org/wiki/Josephus_problem#J > > http://rosettacode.org/wiki/Continued_fraction/Arithmetic/Construct_from_rational_number#J > The continued fraction solution has some oddities. Returning _ to > signal "finished" converts the output vector to floating point. That's > why I excluded the 314285714 100000000x example. The `y is 8' input to > r2cf__CF is arbitrary and I have to strip it with _ from the output > vector. > > ---------------------------------------------------------------------- > For information about J forums see http://www.jsoftware.com/forums.htm > ---------------------------------------------------------------------- For information about J forums see http://www.jsoftware.com/forums.htm
