FW: interesting feature in Dyalog APL/W 
Author Message
 FW: interesting feature in Dyalog APL/W

I the following note there is a referal to "yours truly":
...> in the APL91 paper by Iverson, McDonnell, and
...> yours truly.
But I can not see from the note who "yours truly" is.

Unfortunately I do not have the paper from APL91.
The list of suspects is not very long though.
At least two can be excluded immediately.

It would be nice to know who writes the notes.

Can the system not add that info automatically?
Even if the note is sent directly to the listserv.
There may be people out there who can not guess.
I think credit should go to those who have the credit due.

It looks like it does that sometime and sometime not.
/Gosi

 ----------

Sent:  1. desember 1995 06:13
To:  Multiple recipients of list APL-L
Subject:  Re: interesting feature in Dyalog APL/W

Randy MacDonald:
 The downside is that there is not a big selection of
 function expressions in Dyalog APL, but I AM comparing
 it to J where they can be as elaborate as some actual
 formulas; actually most formulas can be made into
 function expressions.
 Example
      mean =.   +/%#
 which gives you the arithmetic mean of a vector, probably
 cannot be expressed as a function expression in Dyalog.

Bjorn Helgason:
 Why is it that you can do these kinds of things in J and
 only a few in APL ?

 Is the structure of J so much better ?

 If you can do some of the nice things J does why should
 it not be possible to do all of them in APL ?

For the specific question of functional programming aka tacit defn
aka operator expressions, it _is_ possible to add them to APL as
a consistent extension.  Dyalog APL already has one of the needed parts,
viz. function assignment.  The other necessary extensions are:

 - fork
 - atop
 - a way to make a constant function from an array

In particular, you don't need rank.

With these four things, you can write most any explicit function
tacitly; in fact, the translation can be automated (and is  13 : 'blah'
in J).  The almost completeness and the translation from explicit to
tacit are explained in the APL91 paper by Iverson, McDonnell, and
yours truly.  These are consistent extensions in the sense that
they replace what are currently errors in APL.

p.s. Can Dyalog APL gurus please tell me what the dyadic meaning
(if any) of  f{jot}g  is?



Tue, 19 May 1998 03:00:00 GMT  
 
 [ 1 post ] 

 Relevant Pages 

1. interesting feature in Dyalog APL/W

2. Interesting feature in Dyalog APL/W

3. Interesting feature in Dyalog APL/W

4. Trying to hire APL and DYALOG APL for Dallas

5. Converting Dyalog APL Multiple Assignments to APL*PLUS

6. Dyalog APL vs APL*PLUS UNX

7. Freeware Dyalog APL WSS for OO/APL and software management

8. APL*PLUS III vs Dyalog APL/W

9. Fw: Re: ws full

10. MONDAY, 10 OCT: DYALOG APL FOR WINDOWS

11. Dyalog APL/W Version 7.0

12. AP127 for Dyalog APL/W

 

 
Powered by phpBB® Forum Software