# cinloiwt --- get triangle tent weights for 2-D convolution (cinloi()) output  
#
subroutine cinloiwt( a1,a2, lag1,lag2, windwt, w1,w2)
integer 	     a1,a2, lag1,lag2,         w1,w2,     i1,i2, s1,s2, e1,e2
real 				       windwt( w1,w2)
real mid1,mid2, wide1,wide2, x, y
			    call null( windwt, w1*w2)
s1= 1+a1-lag1;	e1= w1-lag1+1;  mid1=(e1+s1)/2.;  wide1=(e1-s1+1.)/2.
s2= 1+a2-lag2;	e2= w2-lag2+1;  mid2=(e2+s2)/2.;  wide2=(e2-s2+1.)/2.

do i1= s1, e1 {	    x = abs((i1-mid1)/wide1)
do i2= s2, e2 {     y = abs((i2-mid2)/wide2)
#	windwt(i1,i2) = amax1( 0., 2. - abs(x+y) - abs(x-y) )		# Cheop
	windwt(i1,i2) = amax1( 0., 1. - abs(x))  * 
			amax1( 0., 1. - abs(y)) 
	}}
return; end