# Commands that generated the results of Section 2.2 of
# "The Portfolio Probe User's Manual"
#
# These commands demonstrate testing the effect of constraint bounds
# using random portfolios
#
# This depends on functions listed in 'pprobe_functions01.R'
# The data are created in 'pprobe_R_data01.txt'
#
# The functions that create the graphs are in 'pprobe_graphFun01.R'
# If you want hardcopies of the graphs, then you need to start a
# graphics device (like 'pdf') and do 'dev.off()' after the graphs
# are created.





# things to CHANGE, possibly

PP_LOC <- ".." 






# running this script to completion requires that
# 'pprobe_R_varCache.txt' has been done
# So test for that:

cat("Testing to see if variance matrices are in place\n",
	"  if you don't get an error from 'attach', it passes\n")
attach(paste(PP_LOC, "/Varshr/varshr.2008-10-14.rda", sep=""))
detach()

# initial setup

if(!exists("us.pricemat")) attach(paste(PP_LOC, "us.pricemat.rda", sep="/"))
if(!exists("us.macdmat")) attach(paste(PP_LOC, "us.macdmat.rda", sep="/"))


# specify quarter end locations
quarter.ends <- c(251, 312, 375, 438, 502, 563, 627, 691, 755)
names(quarter.ends) <- paste(rep(2006:2008, c(1,4,4)), 'Q', c(4,1:3), sep="")

# create a price vector
price.20061229 <- us.pricemat[quarter.ends[1], ]


# create variance matrices
varshr.Q <- vector("list", 8)
names(varshr.Q) <- names(quarter.ends)[1:8]

# calculate return matrix from prices:
us.retmat <- diff(log(us.pricemat))

require(BurStFin)

for(i in 1:8) {
	varshr.Q[[i]] <- var.shrink.eqcor(us.retmat[seq(to=quarter.ends[i]-1, 
		length=250),])
}


# create random portfolios without volatility constraint
rp30w10 <- random.portfolio(number.rand = 10000, prices = price.20061229,  
	gross.value = 1e+06, long.only = TRUE, port.size = 30, 
	max.weight = 0.1)

# create utility over 2007 for fixed portfolios
rp30w10.mvu2.2007 <- pp.meanvarutil(valuation(rp30w10, 
	us.pricemat[(quarter.ends[1]+1):quarter.ends[5],], 
	collapse=TRUE), 2)

# create random portfolios with volatility constraint
rp30w10v08 <- random.portfolio(number.rand = 10000, prices = price.20061229, 
	variance=varshr.Q[[1]],  gross.value = 1e+06, long.only = TRUE, 
	port.size = 30, max.weight = 0.1, var.constraint=.08^2/252)

# create utility for portfolios with volatility constraint
rp30w10v08.mvu2.2007 <- pp.meanvarutil(valuation(rp30w10v08, 
	us.pricemat[(quarter.ends[1]+1):quarter.ends[5],], collapse=TRUE), 2)

# do graph
P.mvu2y2007(NULL)


# This is where the series of cached variances is needed
# see 'pprobe_R_varCache.txt'


# generate rolling random portfolios and calculate utility
# -- constant constraints
rp30w10.mvu2.dat <- pp.date.meanvarutil(rownames(us.pricemat)[251:695], 60, 
	2, us.pricemat, us.macdmat, max.weight=.1, port.size=30, 
	gross.value=1e6, long.only=TRUE, 
	vardir=paste(PP_LOC, "Varshr", sep="/"))

# find approximate minimum variance each day
# we don't care about exact minimum, so restrict computation to 3 iterations
rp30w10.minvar.dat <- pp.date.minvar(rownames(us.pricemat)[251:695], 
	us.pricemat, max.weight=.1, port.size=30, gross.value=1e6, 
	long.only=TRUE, iteration=3, vardir=paste(PP_LOC, "Varshr", sep="/"))

# do graph
P.minvol(NULL)


# generate rolling random portfolios and calculate utility
# -- varialbe volatility constraint
rp30w10vv20.mvu2.dat <- pp.date.var.meanvarutil(
	rownames(us.pricemat)[251:695], 1.44 * rp30w10.minvar.dat, 60, 2,
	us.pricemat, us.macdmat, max.weight=.1, port.size=30, 
	gross.value=1e6, long.only=TRUE, throw.error=FALSE, 
	vardir=paste(PP_LOC, "Varshr", sep="/"))

# do graph
P.mvu2dat(NULL)