No; there is not and cannot ever be a framing error in u&.> because each
box of y is separately opened, operated on, and then enclosed, so that
what is assembled into the final result is always boxes, which are
compatible.
>:&.> 1;'a';3 NB. error in >:
|domain error
| >:&.>1;'a';3
#&.> 1;'a';3 NB. no error in #
+-+-+-+
|1|1|1|
+-+-+-+
Henry Rich
On 3/10/2015 11:57 AM, Don Guinn wrote:
Chapter 4 Section 1 - Diagnosing Errors
The error is reported as a domain error; however, that is incorrect. The
actual error is a framing error in (>). But because the (@) bindes the (>:)
to the (>) The framing error is hidden. I would suggest changing the (>:@>)
to (>:&.>) to avoid the framing error giving an actual domain error. Kind
of interesting in that it shows the first item in the result incremented.
The full statement:
dissect '>:&.> 1;''a'';3'
----------------------------------------------------------------------
For information about J forums see http://www.jsoftware.com/forums.htm
----------------------------------------------------------------------
For information about J forums see http://www.jsoftware.com/forums.htm
