I suspect the original decision, in APL, intended something like this, perhaps along the lines of logical consistency - something like: if an item is not found in a list, it's past the end of it. The number returned in APL was origin-dependent, which made it more of a pain to deal with: I once wrote a cover for dyadic iota to return _1 for "not found" to deal with this. However, since _1 is a valid index in J, this can't be used to indicate "not found".
On Thu, Sep 8, 2011 at 4:44 PM, Steven Taylor <[email protected]> wrote: > ... > I wonder if that was one of the original design intentions of having not > found come back as # rather than _1. > ... > -- Devon McCormick, CFA ^me^ at acm. org is my preferred e-mail ---------------------------------------------------------------------- For information about J forums see http://www.jsoftware.com/forums.htm
