|
Hi Hans-Peter, I tried to check your example with IBM APL2 and it gives me a RANK ERROR on ((∊2=⍴¨x)/,s)⌷x ⍴((∊2=⍴¨x)/,s) is 3 and ⍴x is 4 4. I have earlier noticed some differences in nesting levels between different APL interpreters that could be traced back to what the ISO standard calls "reduction styles" (an implementation parameter). I suppose the reduction style also matters for the EACH operator For example, GNU APL and IBM APL2 use the "Enclose-Reduction-Style" while Dyalog APL uses the "Insert-Reduction-Style". So I wonder where your examples originate. It would help me a lot if you could break down your examples into individual steps like 4⎕CR y1←⍴¯x 4⎕CR y2←2=y1 4⎕CR y3←∈y2 4⎕CR y3/,s ... so that I can see the expected intermediate results (and a hint which ones are wrong) Best Regards, /// Jürgen Sauermann On 11/05/2018 12:49 AM, Hans-Peter
Sorge wrote:
Hi Jürgen, it looks like there is some inconsistency. x←4 4⍴'XX' '' '' '' '' '' '' '' s←(⍳4)∘.,⍳4 s 1 1 1 2 1 3 1 4 2 1 2 2 2 3 2 4 3 1 3 2 3 3 3 4 4 1 4 2 4 3 4 4 ((∊2=⍴¨x)/,s)⌷x XX XX ----- expected (no leading empty elements) XX XX ⍴((∊2=⍴¨x)/,s)⌷x 2 2 ----- expected 2 1 (only one column) ⍴¨((∊2=⍴¨x)/,s)⌷x 0 2 0 2 ----- expected (no leading empty elements) 2 2 ------------------------------------- x←2 2⍴'XX' '' '' '' s←(⍳2)∘.,⍳2 ((∊2=⍴¨x)/,s)⌷x RANK ERROR ((∈2=⍴¨x)/,s)⌷x ^ ^ ------ expected (no RANK ERROR, equivalent result as in previous 4x4 matrix) XX Best Regards Hans-Peter |
signature.asc
Description: OpenPGP digital signature
