I agree, Roger. However, I got the idea from the Henry Rich's Lab "An Introductory Course in J," chapter 10, parts 2 and 3. He did something very elegant called "data-dependent execution," which I was hoping to apply here. Thanks, anyway. Leigh -----Original Message----- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Roger Hui Sent: Tuesday, October 24, 2006 12:55 PM To: Programming forum Subject: Re: [Jprogramming] Choose Operators ^: is not a good way to solve this problem. B, the noun that specifies the selection, is an argument to the operator, not an argument to a verb, and it is awkward to "slice and dice" an operator argument. ----- Original Message ----- From: "Leigh J. Halliwell" <[EMAIL PROTECTED]> Date: Tuesday, October 24, 2006 9:10 am Subject: RE: [Jprogramming] Choose Operators > Thank you, Roger and Cliff. On my first question I was indeed > thinking of > the monadic amend. And Cliff's use of it for my second question > is nice, > too. But I'd still like to know how to make the identity/left > operator work > one-to one, as per my second question: > > "2) X0 and X1 are numeric vectors, and B is a Boolean vector. > They all have > the same length. I'd like to select from X0 where B is 0, and > from X1 where > B is 1. I try the expression: X1 [^:B X0. But the adverb ^:B > wants to get > two-dimensional, i.e., to apply each B to every pair of X1 and X2. > How can > I make J to apply the adverb one-to-one with the arguments?"
---------------------------------------------------------------------- For information about J forums see http://www.jsoftware.com/forums.htm
