function(alpha.level = 1, iota = 0.001, kappa = 0.01, 
	theta = 1.28, start.change = 9, stop.change = 
	10, location = 8)
{
#Program uganda.fit()
#
#iota = 0.001
#kappa = 0.01
#theta = 1.28
#    iota <- x[1]
#    kappa <- x[2]
#    theta <- x[3]
#    start.change <- 9
#    stop.change <- 10
#    location <- 8   #
#
#   #Program for fitting model to Uganda prevalence data by varying
#   #iota, kappa and theta paramters. input.routine.1() will be run
#   #first to set up most objects. This program will overwrite iota.f
#   #iota.m, kappa.f, kappa.m, theta.f and theta.m and then run
#   #project() followed by compute.prevalence() to display graph of
#   #observed and fitted values.
#
#   #overwrite alpha with alpha * alpha.level; this raises hiv mortality
#   #and may enable more reasonable parameter values to give best fit.
#   #Restore original value at end of program to keep it from being mad
#   #successively lower in further runs of this program.
#
	alpha <<- alpha.level * array(data = get.alpha(
		)[1:aa, 1:dd], dim = dim(alpha))	#
#   #overwrite iota, kappa and theta arrays using argument values
#
	iota.f <<- array(data = iota * phi.f, dim = dim(
		iota.f))	#
	iota.m <<- array(data = iota * phi.m, dim = dim(
		iota.m))	#
#
	kappa.f <<- kappa * phi.f	#
	kappa.m <<- kappa * phi.m	#
#
	level.f <- compute.theta.time.trend(
		initial.level = theta, start.change, 
		stop.change)	#
	level.m <- compute.theta.time.trend(
		initial.level = theta, start.change, 
		stop.change)	#
#
	theta.f <<- phi.f %o% level.f	#
	theta.m <<- phi.m %o% level.m	#
#
#   #run project() )
#
	project()	#
#
#   #compute prevalence (see code for compute.prevalence())
#
	num <- apply(zf + zm, c(1, 3), sum)	#
	den <- xf + xm + yf + num	#
	sex <- "Both sexes"	#
	prev <- apply(num[16:50,  ], 2, sum)/apply(den[
		16:50,  ], 2, sum)	#
#
#   #plot model prevalence
#
	times <- 0:(tt - 1)	#
	par(pch = 1)	#
	plot(times, 100 * prev, xlab = "Time", ylab = 
		paste(sex, "prevalence (%)"), ylim = c(
		0, 35))	#
#
#   #Plot observed prevalence, values for 1985-1998 from anoymous unlinked
#   #surveillance of pregnant women. Location allows for changing the year
#   #in which the epidemic begins.
#
	obs.p <- c(10.7, 13.5, 24, 23.5, 24.5, 25, 
		27.82, 29.5, 24.4, 19.2, 18.5, 15.3, 
		14.65, 13.8)	#
	times <- location:(location + length(obs.p) - 1
		)	#
	names(obs.p) <- times	#
	fit.p <- 100 * prev[times + 1]	#
	par(pch = 18)	#
	points(times, obs.p)	#
#
	write.ascii(uganda.fit)
	list(obs.p = obs.p, fit.p = fit.p)
	sum((obs.p - fit.p)^2)
}