although you can write an application to allow your m c n t s is m a d e u n t i l a t s o m e f i n a l s t a g e e x e c u t i o n is

user to highlight text blocks with the mouse, if you demanded. The essence of functional programming
want to do this w h e n using the function editor it's is t h a t f u n c t i o n s do n o t m e r e l y r e t u r n v a l u e s , func-
back to the Control-End key combination. Another t i o n s also a r e v a l u e s , so t h a t f ( x ) - - w h i c h m e a n s t h e
example is the Copy m e n u c o m m a n d which copies r e s u l t o f a p p l y i n g f to x - - i s q u i t e d i f f e r e n t f r o m f
to the Clipboard with Windows, but from wherever itself. T h e i d e a o f a f u n c t i o n i t s e l f b e i n g a v a l u e is
tagged lines are held in the API.JW function editor. r e f l e c t e d in A P L 2 o p e r a t o r s , a n d t h e p r e s e n c e o f
s a y i n n e r p r o d u c t in t h e l a n g u a g e is a c l e a r
a c k n o w l e d g m e n t t h a t it is p o s s i b l e t o p e r f o r m t h e
The Future f u n c t i o n a l g e b r a f.g r e g a r d l e s s o f a n y d a t a to w h i c h
D e s p i t e all o f t h e s e f r u s t r a t i o n s , A P L / W c o n t i n - t h i s f u n c t i o n c o m b i n a t i o n m i g h t u l t i m a t e l y be
u e s to be m y A P L o f p r e f e r e n c e . I c a n s y m p a t h i s e applied. S i n c e A P L 2 a l l o w s o p e r a t o r s to be user-de-
w i t h t h e d e c i s i o n s t h a t D y a d i c a r e h a v i n g to m a k e fined, t h e s c o p e for f u n c t i o n a l p r o g r a m m i n g in t h i s
in o r d e r to b r i n g a u s e a b l e p r o d u c t t o m a r k e t ; a n d s e n s e is u n l i m i t e d .
j u s t h o p e t h a t t h e r e s u l t i n g r e v e n u e s t r e a m is g o i n g As a s i m p l e i l l u s t r a t i o n h e r e is a n e x a m p l e o f
to be l a r g e e n o u g h to p e r m i t t h e s h o r t c o m i n g s to t w o o p e r a t i o n s w h i c h a r e i d e n t i c a l in t h e i r p r o g r a m
be a d d r e s s e d . As t h o s e o f y o u w h o w e r e p r e s e n t a t bodies, but which are quite different at a functional
D y a d i c ' s A P L 9 3 V e n d o r F o r u m will k n o w . t h e com- p r o g r a m m i n g level.
p a n y h a s s o m e v e r y a t t r a c t i v e p l a n s for t h e f u t u r e .
W e n o w h a v e t h r e e s i g n i f i c a n t p l a y e r s in t h e Z÷L MULl R Z÷(L MUL2)R
" A P L for W i n d o w s " a r e n a , all t a k i n g t h e i r d i f f e r e n t Z÷LxR Z÷LxR
stances. I h a v e n ' t b e e n a b l e to t a k e a n y s e r i o u s M U L l has two arguments, M U L 2 has only one.
look at Manugistics most recent versions of 2 M U L 2 3 means apply f = 2 MUL2--ca]I this the
APL*PLUS II, a n d I d o n ' t r e l i s h t h e a m o u n t o f " d o u b l e " f u n c t i o n - - t o a r g u m e n t 3 to r e t u r n 6. I t
s t u d y t h a t s e e m s to be n e c e s s a r y to c o m e to g r i p s is m e a n i n g f u l to t a l k a b o u t f 2 (3), w h i c h h a s t h e
w i t h t h e i r i n t e r f a c e ( a l t h o u g h I ' d h a p p i l y be p r o v e d v a l u e 1 2-
mistaken). I've looked briefly at ISI's APLIWIN, D y a d i c o p e r a t o r s c o m b i n e f u n c t i o n s in t h e s a m e
and see a product which combines technical virtue way that dyadic functions combine data. For exam-
with an individually-justifiable price (and noted that ple e l s e is a r e a d i l y c o n c e i v a b l e f u n c t i o n o f func-
D D E b e t w e e n A P L I W I N a n d A c c e s s a p p e a r s to t i o n s for w h i c h e l s e ~ g ) m e a n s "f, or i f n o t E-"
w o r k j u s t fine). A P L 2 for O8/2 b e c k o n s f r o m t h e T h e c o n d i t i o n s in e x e c u t i o n u n d e r w h i c h e i t h e r f or
wings to take us into that particular arena. g operates are a matter of data, not of function
O n e of the things that was most rewarding a l g e b r a . T h e l a t t e r is c o n c e r n e d o n l y w i t h d e f i n i n g
about A P L 9 3 was the very positive feeling which a f u n c t i o n w h i c h c o m b i n e s f a n d E, a n d in w h i c h f
people were beginning to be able to generate about is d o m i n a n t in s o m e d a t a - d r i v e n s e n s e . I m p l e m e n t a -
the future for APL; it was more than wishful think- tion of ELSE begins with two program body lines
ing. To paraphrase what another vendor said not
so long ago--we n o w have some excellent A P L Z+(P ELSE Q )R
development tools; the next step is fox us to start Z÷P R
building great applications with them. m Z÷Q R
both of which must be present but only one of
which is selected depending on data. At the second
stage of development a predicate L might be intro-
Functional Programming with APL2 duced to settle the issue of P or Q:
Z÷L(P ELSE Q )R
÷L 1 IF~L
Norman Thomson ÷0 Z÷P R
Mail Point 17H
LI:Z÷Q R
I B M U K Ltd.
P O B o x 30, (X~3)( ÷ELSE+ )X+2
G r e e n e ck, 0.25
Scotland PAl5 6AP (X~3)( ÷ELSE+ )X÷3
Tel: +44-475-895165 3
E-Maih nthomson@GNKVM2. VNET. IBM. com
A m o r e i n t e r e s t i n g i l l u s t r a t i o n is in t h e field o f
Functional programming is a style of program- iterative solution of numerical equations where the
ming which recognises that functions per se obey o p e r a t o r U N T I L h a s a f u n d a m e n t a l r o l e in t h e r e l e -
mathematical principles, quite separately from their v a n t f u n c t i o n algebra_
contents. In its purest form no reference to argu-

APL Q~ote ~ a d 15
 Z÷(P UNTIL Q)R F NEWTON -1.2
÷LI IF Q R -0.9999999999
÷0 Z÷(P UNTIL Q)P R F NEWTON 1
L1 :Z÷R 2.000000001
U N T I L is a c o n n e c t i n g o p e r a t i o n b e t w e e n f u n c t i o n s A n a l t e r n a t i v e root-finding a l g o r i t h m to N e w t o n -
in m u c h the same w a y as addi t i on or m u l t i p l i c a t i o n Raphson is t h e so-called S e c a n t M e t h o d . Again
a r e o p e r a t i o n s w h i c h c o m b i n e n u m b e r s in simple U N T I L act s as a c o n n e c t o r b e t w e e n a step f u n c t i o n
arithmetic. In in f or m al t e r m s t h e above o p e r a t o r and a stopping c r i t e r i o n .
definition says t h a t t h e f u n c t i o n c o m b i n a t i o n
P UNTIL Q (the d e r i v e d f u n c t i o n in APL2 t e r m s ) Z÷(P SECANT )START
is i n v a r i a n t u n d e r t h e o p e r a t i o n o f P, s u b j e c t to t h e Z÷÷(P SECSTEP UNTIL(P EQUAL) )START
p r e d i c a t e Q r e m a i n i n g false.
As an example c o n s i d e r t h e following simple Z÷(P SECSTEP)X;T
i t e r a t i v e a l g o r i t h m for e x t r a c t i n g a sqt~are root. Z÷TpX P OPPOSITE(-/(4pX)xT)÷-IT÷P"X
T h e q u a n t i t y E P S should be u n d e r s t o o d as defining
t h e d i f f e r e n c e b e t w e e n s u c c e s s i v e i t e r a t e s w h i c h is Z÷L(P OPPOSITE)R;T
c o n s i d e r e d small e n o u g h to stop the al gor i t h m . Z÷(¢( xP R)=xP"T)/T÷(÷L),R,I+L
Z÷START SQRT Y
Z÷(P EQUAL)X;T
Z÷(IMPROVE UNTIL NEAR )START
Z ÷ E P S > IP ÷ X
Z÷IMPROVE X Z÷NEAR X
F SECANT I 3
Z÷O.5xX+Y÷X Z ÷ E P S > IY-X*2
1.99999857
S Q R T can be thought of as defining the procesa of
In a p s e u d o - m a t h e m a t i c a l not at i on:
extracting a square root using the particular iter-
ation sequence whose step is defined by .improve, secant(f) = Until(secstep(f),equal(f))
and whose stopping criterion by near.
In order to describe another iteration process, Another iterative method for finding the roots of f
for example Newton-Raphson iteration, it is neces- consists o f r e - e x p r e s s i n g f ( x ) = 0 as x = g ( x ) and
sary only to identify the analogous step function t h e n c a r r y i n g o u t r e p e a t e d s u b s t i t u t i o n in the for-
and stopping criterion: mula:

Z÷( P N E W T O N ) S T A R T Xr + 1 = g(Xr)
Z÷(P NEWSTEP UNTIL(P EQUAL) )START T h e same f ( x ) is used as above, r e c o g n i z i n g t h a t
t h e r e are two possible forms for g(x), n a m e l y x 2 - 2
Z÷(P NEWSTEP)X Z÷(P EQUAL)X;T and ~/~ + 2 -
Z÷X-(P X)÷P DERIV X Z+EPS>IP X Z÷(P ITER)START
Z÷(P UNTIL(P THERE) )START
Z÷(P DERIV)X
Z÷( (P X + A X ) - P X)÷AX Z÷GI X Z÷(P THERE)X
Functional programming style leads to a clear sepa- Z÷-2+X*2 Z ÷ E P S > IX-P X
ration between process description, and process
instance, which is what is more usually called exe- Z÷G2 X
cution. Z ÷ ( X + 2 )* .5
As an illustration of the latter, let f(x) be the
f u n c t i o n 2 - - x ( x - 1). To solve f ( x ) = 0 define t h e G1 ITER 1 G2 ITER 1
APL2 f u n c t i o n F: -1 1.999990588
Z÷F X Thes~uct~e
Z÷2-XxX-I
iterative_method (f) =Until (step(f), stop (f))
and s e t
is v e r y general , and e m e r g e s from t a k i n g a func-
6X tional p r o g r a m m i n g v i e w p o i n t o f t h e problem, o r
0.00005 r a t h e r family of problems. It is hoped t h a t t h e
T h e n use the function F as the left; argument i l l u s t r a t i o n s in this art i cl e show t h a t a f u n c t i o n a l
of
N E W T O N to obtain roots: p r o g r a m m i n g a p p r o a c h to problem-solving c a n in
t he r i g h t c i r c u m s t a n c e s lead to v e r y s t r i k i n g uses
o f t h e APL2 o p e r a t o r facility. •

16 December 1993 - Volume 24, Number 2