JPhrases/LinearVectorFunc

9B. Linear Vector Functions

If f(x+y)equals (f x)+(f y) for all vectors x and y in its domain, then f is said to be a linear vector function. In other words, f@:+ -: f@[ + f@] is a test for the linearity of f:

   f=: +/\"1	  NB. Sum scan (subtotals) is a linear vector function
   x=: 4 3 2 1 0
   y=: 2 3 5 7 11
   x ([ , ] , f@[ , f@] , (f@[ + f@]) ,: f@:+) y
4  3  2  1  0	x
2  3  5  7 11	y
4  7  9 10 10	f x
2  5 10 17 28	f y
6 12 19 27 38	(f x)+(f y)
6 12 19 27 38	f (x+y)

For vector arguments of dimension n, any linear vector function f can be expressed as a matrix product, using a matrix obtained by applying f to the identity matrix of order n. Moreover, the inverse of a linear function is a linear function. For example:

   g=: f^:_1	    NB. Inverse of subtotals is first differences
   (] ; f ; g ; f@g) y
+--------------------------------------------+
|2 3 5 7 11|2 5 10 17 28|2 1 2 2 4|2 3 5 7 11|
+--------------------------------------------+

   I=: = i. 5	    NB. Identity matrix
   mf=: (f I) [ mg=: (g I)
   mp=: +/ . *	    NB. Matrix product
   mf ; mg ; (y,(y mp mf),:f y) ; (mf mp mg)
+----------------------------------------------+
|1 1 1 1 1|1 _1  0  0  0|2 3  5  7 11|1 0 0 0 0|
|0 1 1 1 1|0  1 _1  0  0|2 5 10 17 28|0 1 0 0 0|
|0 0 1 1 1|0  0  1 _1  0|2 5 10 17 28|0 0 1 0 0|
|0 0 0 1 1|0  0  0  1 _1|            |0 0 0 1 0|
|0 0 0 0 1|0  0  0  0  1|            |0 0 0 0 1|
+----------------------------------------------+

a0=: MR=: 1 : 'u@=@i.@#'	Matrix representation of linear function
d1=: mp=: +/ . *	Matrix product
a2=: L=: &mp	Linear function represented by matrix
a3=: inv=: ^:_1	Inverse adverb
a4=: MRI=: (^:_1) 'MR' f.	Matrix representation of inverse function
a5=: MR (%.@)	"

   m=: f MR y [ mi=: f MRI y
   m ; mi ; y,(y L m),:(mi L m L y)
+------------------------------------+
|1 1 1 1 1|1 _1  0  0  0|2 3  5  7 11|
|0 1 1 1 1|0  1 _1  0  0|2 5 10 17 28|
|0 0 1 1 1|0  0  1 _1  0|2 5 10 17 28|
|0 0 0 1 1|0  0  0  1 _1|            |
|0 0 0 0 1|0  0  0  0  1|            |
+------------------------------------+

---