######################################################
# Stat 6721 - Assignment 3
#
# Sarah Rathwell
#
# Objective - Investigate Merck and Pfizer for 2014
# and trade on 2015 - Model selection.
######################################################
rm(list=ls())
library(quantmod)
## Loading required package: xts
## Loading required package: zoo
##
## Attaching package: 'zoo'
## The following objects are masked from 'package:base':
##
## as.Date, as.Date.numeric
## Loading required package: TTR
## Version 0.4-0 included new data defaults. See ?getSymbols.
######################################################
cat(" Third look - Merk vs Pfizer.\n\n",
"Last modification:",date(),'\n')
## Third look - Merk vs Pfizer.
##
## Last modification: Fri Nov 18 11:43:50 2016
######################################################
MERK <- getSymbols('MRK', from = '2010-01-01',to = '2014-12-31',
adjust = T, auto.assign = FALSE)
## As of 0.4-0, 'getSymbols' uses env=parent.frame() and
## auto.assign=TRUE by default.
##
## This behavior will be phased out in 0.5-0 when the call will
## default to use auto.assign=FALSE. getOption("getSymbols.env") and
## getOptions("getSymbols.auto.assign") are now checked for alternate defaults
##
## This message is shown once per session and may be disabled by setting
## options("getSymbols.warning4.0"=FALSE). See ?getSymbols for more details.
PFE <- getSymbols('PFE', from = '2010-01-01',to = '2014-12-31',
adjust = T, auto.assign = FALSE)
head(MERK)
## MRK.Open MRK.High MRK.Low MRK.Close MRK.Volume MRK.Adjusted
## 2010-01-04 30.24979 30.60306 30.02797 30.40589 13896500 28.72234
## 2010-01-05 30.64414 30.76738 30.34016 30.52912 14744800 28.83875
## 2010-01-06 30.57020 30.99741 30.31552 30.93990 15239900 29.22679
## 2010-01-07 30.78381 31.13708 30.72630 30.98920 11916600 29.27335
## 2010-01-08 31.11243 31.12886 30.75094 30.97277 10767100 29.25783
## 2010-01-11 31.12065 31.15351 30.80845 31.09600 9581800 29.37424
head(PFE)
## PFE.Open PFE.High PFE.Low PFE.Close PFE.Volume PFE.Adjusted
## 2010-01-04 15.13582 15.69088 15.11097 15.68260 52086000 14.62283
## 2010-01-05 15.67431 15.68260 15.36778 15.45892 43372800 14.41426
## 2010-01-06 15.45892 15.58318 15.33465 15.40921 41405100 14.36792
## 2010-01-07 15.44235 15.46720 15.29322 15.35122 39428000 14.31384
## 2010-01-08 15.42578 15.50034 15.34293 15.47548 30407700 14.42971
## 2010-01-11 15.59975 15.69917 15.46720 15.59975 32447500 14.54558
MRK.a <- MERK[,6]
PFE.a <- PFE[,6]
MRK.ld <- diff(log(MRK.a))
MRK.ld <- MRK.ld[-1,]
PFE.ld <- diff(log(PFE.a))
PFE.ld <- PFE.ld[-1,]
par(mfrow=c(1,2),oma = c(0, 0, 2, 0))
acf(MRK.ld, main="")
pacf(MRK.ld, main="")
mtext('Correlation Functions: MRK', outer = TRUE, cex = 1.5)
acf(PFE.ld, main="")
pacf(PFE.ld, main="")
mtext('Correlation Functions: PFE', outer = TRUE, cex = 1.5)
######################################################
tsdiag.new <- function (object, gof.lag = 10, ...)
{
rs <- object$residuals
stdres <- rs/sqrt(object$sigma2)
nlag <- gof.lag
pval <- numeric(nlag)
for (i in 1:nlag) pval[i] <- Box.test(rs, i, type = "Ljung-Box")$p.value
plot(1:nlag, pval, xlab = "lag", ylab = "p value", ylim = c(0, 1),
main = "p values for Ljung-Box statistic", sub = object$call$order)
abline(h = 0.05, lty = 2, col = "blue")
}
par(mfrow=c(1,1), oma=c(0,0,0,0))
PFE.arma11 <- arima(PFE.ld, order=c(1,0,1))
PFE.arma11$code
## [1] 0
PFE.arma11
##
## Call:
## arima(x = PFE.ld, order = c(1, 0, 1))
##
## Coefficients:
## ar1 ma1 intercept
## -0.0335 -0.0294 5e-04
## s.e. 0.4447 0.4427 3e-04
##
## sigma^2 estimated as 0.0001357: log likelihood = 3813.2, aic = -7618.4
tsdiag.new(PFE.arma11)
PFE.arma10 <- arima(PFE.ld, order=c(1,0,0))
PFE.arma10$code
## [1] 0
PFE.arma10
##
## Call:
## arima(x = PFE.ld, order = c(1, 0, 0))
##
## Coefficients:
## ar1 intercept
## -0.0630 5e-04
## s.e. 0.0282 3e-04
##
## sigma^2 estimated as 0.0001357: log likelihood = 3813.21, aic = -7620.41
tsdiag.new(PFE.arma10)
PFE.arma01 <- arima(PFE.ld, order=c(0,0,1))
PFE.arma01$code
## [1] 0
PFE.arma01
##
## Call:
## arima(x = PFE.ld, order = c(0, 0, 1))
##
## Coefficients:
## ma1 intercept
## -0.0624 5e-04
## s.e. 0.0280 3e-04
##
## sigma^2 estimated as 0.0001357: log likelihood = 3813.18, aic = -7620.37
tsdiag.new(PFE.arma01)
# no
PFE.arma00 <- arima(PFE.ld, order=c(0,0,0))
PFE.arma00$code
## [1] 0
PFE.arma00
##
## Call:
## arima(x = PFE.ld, order = c(0, 0, 0))
##
## Coefficients:
## intercept
## 5e-04
## s.e. 3e-04
##
## sigma^2 estimated as 0.0001362: log likelihood = 3810.71, aic = -7617.42
tsdiag.new(PFE.arma00)
# no
PFE.armalist <- list(PFE.arma11, PFE.arma10, PFE.arma01)
# lowest are 01, 10
# --> 1 parameter
MRK.arma55 <- arima(MRK.ld, order=c(5,0,5))
MRK.arma55$code
## [1] 0
MRK.arma55
##
## Call:
## arima(x = MRK.ld, order = c(5, 0, 5))
##
## Coefficients:
## ar1 ar2 ar3 ar4 ar5 ma1 ma2 ma3
## -0.1675 -0.3450 -0.1551 -0.2967 -0.8504 0.1688 0.3936 0.1108
## s.e. 0.0895 0.0644 0.0809 0.0831 0.1060 0.1121 0.0887 0.0903
## ma4 ma5 intercept
## 0.3353 0.8066 5e-04
## s.e. 0.0914 0.1202 3e-04
##
## sigma^2 estimated as 0.0001383: log likelihood = 3798.96, aic = -7573.92
tsdiag.new(MRK.arma55)
MRK.arma54 <- arima(MRK.ld, order=c(5,0,4))
MRK.arma54$code
## [1] 0
MRK.arma54
##
## Call:
## arima(x = MRK.ld, order = c(5, 0, 4))
##
## Coefficients:
## ar1 ar2 ar3 ar4 ar5 ma1 ma2 ma3
## 0.5156 0.0734 -0.6255 0.0623 -0.0873 -0.5422 -0.0507 0.5864
## s.e. 0.3831 0.2355 0.2495 0.3401 0.0369 0.3844 0.2423 0.2479
## ma4 intercept
## -0.0383 5e-04
## s.e. 0.3356 3e-04
##
## sigma^2 estimated as 0.0001401: log likelihood = 3793.33, aic = -7564.67
tsdiag.new(MRK.arma54)
MRK.arma53 <- arima(MRK.ld, order=c(5,0,3))
MRK.arma53$code
## [1] 0
MRK.arma53
##
## Call:
## arima(x = MRK.ld, order = c(5, 0, 3))
##
## Coefficients:
## ar1 ar2 ar3 ar4 ar5 ma1 ma2 ma3
## 0.4770 0.0833 -0.6049 0.0239 -0.0889 -0.5034 -0.0614 0.5673
## s.e. 0.1707 0.2178 0.1573 0.0401 0.0332 0.1705 0.2211 0.1701
## intercept
## 5e-04
## s.e. 3e-04
##
## sigma^2 estimated as 0.0001401: log likelihood = 3793.32, aic = -7566.64
tsdiag.new(MRK.arma53)
MRK.arma52 <- arima(MRK.ld, order=c(5,0,2))
MRK.arma52$code
## [1] 0
MRK.arma52
##
## Call:
## arima(x = MRK.ld, order = c(5, 0, 2))
##
## Coefficients:
## ar1 ar2 ar3 ar4 ar5 ma1 ma2
## 0.1965 -0.0743 -0.0492 0.0268 -0.1002 -0.2206 0.0899
## s.e. 0.3275 0.2220 0.0300 0.0319 0.0292 0.3296 0.2266
## intercept
## 5e-04
## s.e. 3e-04
##
## sigma^2 estimated as 0.0001407: log likelihood = 3790.28, aic = -7562.55
tsdiag.new(MRK.arma52)
MRK.arma51 <- arima(MRK.ld, order=c(5,0,1))
MRK.arma51$code
## [1] 0
MRK.arma51
##
## Call:
## arima(x = MRK.ld, order = c(5, 0, 1))
##
## Coefficients:
## ar1 ar2 ar3 ar4 ar5 ma1 intercept
## 0.1419 0.0133 -0.0462 0.0232 -0.0958 -0.1652 5e-04
## s.e. 0.2450 0.0290 0.0285 0.0304 0.0281 0.2458 3e-04
##
## sigma^2 estimated as 0.0001408: log likelihood = 3790.2, aic = -7564.4
tsdiag.new(MRK.arma51)
MRK.arma50 <- arima(MRK.ld, order=c(5,0,0))
MRK.arma50$code
## [1] 0
MRK.arma50
##
## Call:
## arima(x = MRK.ld, order = c(5, 0, 0))
##
## Coefficients:
## ar1 ar2 ar3 ar4 ar5 intercept
## -0.0219 0.0095 -0.0439 0.0158 -0.0927 5e-04
## s.e. 0.0281 0.0281 0.0281 0.0281 0.0281 3e-04
##
## sigma^2 estimated as 0.0001408: log likelihood = 3790.03, aic = -7566.06
tsdiag.new(MRK.arma50)
MRK.arma45 <- arima(MRK.ld, order=c(4,0,5))
MRK.arma45$code
## [1] 0
MRK.arma45
##
## Call:
## arima(x = MRK.ld, order = c(4, 0, 5))
##
## Coefficients:
## ar1 ar2 ar3 ar4 ma1 ma2 ma3 ma4
## 0.4732 0.1835 -0.6545 -0.0071 -0.5001 -0.1612 0.6190 0.0245
## s.e. 0.3405 0.2256 0.2036 0.2916 0.3393 0.2328 0.2022 0.2834
## ma5 intercept
## -0.0833 5e-04
## s.e. 0.0338 3e-04
##
## sigma^2 estimated as 0.0001401: log likelihood = 3793.28, aic = -7564.57
tsdiag.new(MRK.arma45)
MRK.arma44 <- arima(MRK.ld, order=c(4,0,4))
MRK.arma44$code
## [1] 0
MRK.arma44
##
## Call:
## arima(x = MRK.ld, order = c(4, 0, 4))
##
## Coefficients:
## Warning in sqrt(diag(x$var.coef)): NaNs produced
## ar1 ar2 ar3 ar4 ma1 ma2 ma3 ma4
## -0.4637 0.7527 0.5195 0.1075 0.4415 -0.7568 -0.5460 -0.095
## s.e. 0.9901 NaN 0.8625 NaN 0.9924 NaN 0.8532 NaN
## intercept
## 5e-04
## s.e. 2e-04
##
## sigma^2 estimated as 0.0001411: log likelihood = 3788.7, aic = -7557.41
tsdiag.new(MRK.arma44)
MRK.arma43 <- arima(MRK.ld, order=c(4,0,3))
MRK.arma43$code
## [1] 0
MRK.arma43
##
## Call:
## arima(x = MRK.ld, order = c(4, 0, 3))
##
## Coefficients:
## Warning in sqrt(diag(x$var.coef)): NaNs produced
## ar1 ar2 ar3 ar4 ma1 ma2 ma3 intercept
## -0.5576 0.8462 0.6204 0.0104 0.5356 -0.8522 -0.6421 5e-04
## s.e. NaN 0.0785 NaN 0.0335 NaN 0.0658 NaN 2e-04
##
## sigma^2 estimated as 0.0001411: log likelihood = 3788.7, aic = -7559.4
tsdiag.new(MRK.arma43)
MRK.arma42 <- arima(MRK.ld, order=c(4,0,2))
MRK.arma42$code
## [1] 0
MRK.arma42
##
## Call:
## arima(x = MRK.ld, order = c(4, 0, 2))
##
## Coefficients:
## ar1 ar2 ar3 ar4 ma1 ma2 intercept
## -0.1366 0.5883 -0.0475 -0.0014 0.1157 -0.5820 5e-04
## s.e. 0.3112 0.2282 0.0317 0.0388 0.3099 0.2229 3e-04
##
## sigma^2 estimated as 0.0001413: log likelihood = 3787.59, aic = -7559.17
tsdiag.new(MRK.arma42)
# no
MRK.arma41 <- arima(MRK.ld, order=c(4,0,1))
MRK.arma41$code
## [1] 0
MRK.arma41
##
## Call:
## arima(x = MRK.ld, order = c(4, 0, 1))
##
## Coefficients:
## ar1 ar2 ar3 ar4 ma1 intercept
## -0.7816 -0.0052 -0.0340 0.0063 0.7622 5e-04
## s.e. 0.1912 0.0361 0.0359 0.0357 0.1893 3e-04
##
## sigma^2 estimated as 0.0001415: log likelihood = 3786.9, aic = -7559.8
tsdiag.new(MRK.arma41)
# no
MRK.arma35 <- arima(MRK.ld, order=c(3,0,5))
MRK.arma35$code
## [1] 0
MRK.arma35
##
## Call:
## arima(x = MRK.ld, order = c(3, 0, 5))
##
## Coefficients:
## ar1 ar2 ar3 ma1 ma2 ma3 ma4 ma5
## 0.4952 0.1591 -0.6458 -0.5222 -0.1364 0.6080 0.0202 -0.0840
## s.e. 0.1668 0.2111 0.1518 0.1676 0.2155 0.1654 0.0395 0.0316
## intercept
## 5e-04
## s.e. 3e-04
##
## sigma^2 estimated as 0.0001401: log likelihood = 3793.28, aic = -7566.57
tsdiag.new(MRK.arma35)
MRK.arma34 <- arima(MRK.ld, order=c(3,0,4))
MRK.arma34$code
## [1] 0
MRK.arma34
##
## Call:
## arima(x = MRK.ld, order = c(3, 0, 4))
##
## Coefficients:
## Warning in sqrt(diag(x$var.coef)): NaNs produced
## ar1 ar2 ar3 ma1 ma2 ma3 ma4 intercept
## -0.2274 0.6358 0.0973 0.2069 -0.6353 -0.1408 0.0062 5e-04
## s.e. NaN NaN NaN NaN NaN NaN NaN 3e-04
##
## sigma^2 estimated as 0.0001413: log likelihood = 3787.6, aic = -7557.19
tsdiag.new(MRK.arma34)
MRK.arma33 <- arima(MRK.ld, order=c(3,0,3))
MRK.arma33$code
## [1] 0
MRK.arma33
##
## Call:
## arima(x = MRK.ld, order = c(3, 0, 3))
##
## Coefficients:
## Warning in sqrt(diag(x$var.coef)): NaNs produced
## ar1 ar2 ar3 ma1 ma2 ma3 intercept
## -0.0851 0.5947 -0.0999 0.0641 -0.5879 0.0531 5e-04
## s.e. NaN 0.1988 NaN NaN 0.1949 NaN 3e-04
##
## sigma^2 estimated as 0.0001413: log likelihood = 3787.6, aic = -7559.19
tsdiag.new(MRK.arma33)
MRK.arma32 <- arima(MRK.ld, order=c(3,0,2))
MRK.arma32$code
## [1] 0
MRK.arma32
##
## Call:
## arima(x = MRK.ld, order = c(3, 0, 2))
##
## Coefficients:
## ar1 ar2 ar3 ma1 ma2 intercept
## -0.1002 0.6081 -0.0473 0.0793 -0.6004 5e-04
## s.e. 0.2408 0.2291 0.0323 0.2398 0.2197 3e-04
##
## sigma^2 estimated as 0.0001413: log likelihood = 3787.57, aic = -7561.15
tsdiag.new(MRK.arma32)
#no
MRK.arma31 <- arima(MRK.ld, order=c(3,0,1))
MRK.arma31$code
## [1] 0
MRK.arma31
##
## Call:
## arima(x = MRK.ld, order = c(3, 0, 1))
##
## Coefficients:
## ar1 ar2 ar3 ma1 intercept
## -0.8029 -0.0057 -0.0368 0.7834 5e-04
## s.e. 0.1423 0.0363 0.0321 0.1401 3e-04
##
## sigma^2 estimated as 0.0001415: log likelihood = 3786.9, aic = -7561.79
tsdiag.new(MRK.arma31)
#no
MRK.arma25 <- arima(MRK.ld, order=c(2,0,5))
MRK.arma25$code
## [1] 0
MRK.arma25
##
## Call:
## arima(x = MRK.ld, order = c(2, 0, 5))
##
## Coefficients:
## ar1 ar2 ma1 ma2 ma3 ma4 ma5
## 0.2138 -0.0993 -0.2375 0.1157 -0.0462 0.0298 -0.0854
## s.e. 0.4035 0.2610 0.4025 0.2651 0.0311 0.0359 0.0278
## intercept
## 5e-04
## s.e. 3e-04
##
## sigma^2 estimated as 0.000141: log likelihood = 3789.17, aic = -7560.34
tsdiag.new(MRK.arma25)
MRK.arma24 <- arima(MRK.ld, order=c(2,0,4))
MRK.arma24$code
## [1] 0
MRK.arma24
##
## Call:
## arima(x = MRK.ld, order = c(2, 0, 4))
##
## Coefficients:
## Warning in sqrt(diag(x$var.coef)): NaNs produced
## ar1 ar2 ma1 ma2 ma3 ma4 intercept
## -0.0599 0.691 0.0394 -0.6849 -0.0423 0.0101 5e-04
## s.e. NaN NaN NaN NaN 0.0300 0.0273 3e-04
##
## sigma^2 estimated as 0.0001414: log likelihood = 3787.55, aic = -7559.1
tsdiag.new(MRK.arma24)
MRK.arma23 <- arima(MRK.ld, order=c(2,0,3))
MRK.arma23$code
## [1] 0
MRK.arma23
##
## Call:
## arima(x = MRK.ld, order = c(2, 0, 3))
##
## Coefficients:
## ar1 ar2 ma1 ma2 ma3 intercept
## -0.1677 0.6027 0.1469 -0.5978 -0.0462 5e-04
## s.e. 0.2733 0.2598 0.2739 0.2473 0.0323 3e-04
##
## sigma^2 estimated as 0.0001413: log likelihood = 3787.58, aic = -7561.16
tsdiag.new(MRK.arma23)
#no
MRK.arma22 <- arima(MRK.ld, order=c(2,0,2))
MRK.arma22$code
## [1] 0
MRK.arma22
##
## Call:
## arima(x = MRK.ld, order = c(2, 0, 2))
##
## Coefficients:
## ar1 ar2 ma1 ma2 intercept
## 0.0774 0.8800 -0.1202 -0.8585 5e-04
## s.e. 0.0659 0.0632 0.0722 0.0706 2e-04
##
## sigma^2 estimated as 0.0001413: log likelihood = 3787.88, aic = -7563.77
tsdiag.new(MRK.arma22)
#no
MRK.arma15 <- arima(MRK.ld, order=c(1,0,5))
MRK.arma15$code
## [1] 0
MRK.arma15
##
## Call:
## arima(x = MRK.ld, order = c(1, 0, 5))
##
## Coefficients:
## ar1 ma1 ma2 ma3 ma4 ma5 intercept
## 0.1477 -0.1714 0.0154 -0.0421 0.0249 -0.0812 5e-04
## s.e. 0.3093 0.3084 0.0299 0.0289 0.0322 0.0262 3e-04
##
## sigma^2 estimated as 0.000141: log likelihood = 3789.1, aic = -7562.21
tsdiag.new(MRK.arma15)
#no
MRK.arma14 <- arima(MRK.ld, order=c(1,0,4))
MRK.arma14$code
## [1] 0
MRK.arma14
##
## Call:
## arima(x = MRK.ld, order = c(1, 0, 4))
##
## Coefficients:
## ar1 ma1 ma2 ma3 ma4 intercept
## -0.8405 0.8209 -0.0086 -0.0388 -0.0012 5e-04
## s.e. 0.1451 0.1467 0.0377 0.0404 0.0357 3e-04
##
## sigma^2 estimated as 0.0001415: log likelihood = 3786.96, aic = -7559.92
tsdiag.new(MRK.arma14)
#no
MRK.arma05 <- arima(MRK.ld, order=c(0,0,5))
MRK.arma05$code
## [1] 0
MRK.arma05
##
## Call:
## arima(x = MRK.ld, order = c(0, 0, 5))
##
## Coefficients:
## ma1 ma2 ma3 ma4 ma5 intercept
## -0.0246 0.0111 -0.0408 0.0170 -0.0785 5e-04
## s.e. 0.0282 0.0282 0.0286 0.0273 0.0259 3e-04
##
## sigma^2 estimated as 0.000141: log likelihood = 3789, aic = -7564.01
tsdiag.new(MRK.arma05)
#no
MRK.arma04 <- arima(MRK.ld, order=c(0,0,4))
MRK.arma04$code
## [1] 0
MRK.arma04
##
## Call:
## arima(x = MRK.ld, order = c(0, 0, 4))
##
## Coefficients:
## ma1 ma2 ma3 ma4 intercept
## -0.0211 0.0064 -0.0463 0.0159 5e-04
## s.e. 0.0285 0.0286 0.0291 0.0275 3e-04
##
## sigma^2 estimated as 0.0001421: log likelihood = 3784.45, aic = -7556.89
tsdiag.new(MRK.arma04)
#no
MRK.armalist <- list(MRK.arma55, MRK.arma54, MRK.arma53, MRK.arma52, MRK.arma51,
MRK.arma50, MRK.arma45, MRK.arma44, MRK.arma43, MRK.arma42,
MRK.arma35, MRK.arma34, MRK.arma33, MRK.arma32, MRK.arma25,
MRK.arma24, MRK.arma23, MRK.arma15, MRK.arma05)
# lowest are 05, 23, 32, 50
# --> 5 parameters
aiccall.list <- function(armalist){
test.aic <- NA
test.call <- NA
test.code <- NA
for (i in 1:length(armalist)){
test.aic[i] <- round(armalist[[i]]$aic, digits = 3)
test.call[i] <- as.character(armalist[[i]]$call[3])
test.code[i] <- armalist[[i]]$code
}
output <- as.data.frame(cbind(test.aic, test.call, test.code))
colnames(output) <- c("AIC", "Order", "Code")
return(output)
}
PFE.final <- aiccall.list(PFE.armalist)
PFE.final
## AIC Order Code
## 1 -7618.398 c(1, 0, 1) 0
## 2 -7620.414 c(1, 0, 0) 0
## 3 -7620.369 c(0, 0, 1) 0
PFE.final[PFE.final$Code==1,] #no convergence issues
## [1] AIC Order Code
## <0 rows> (or 0-length row.names)
range(as.numeric(levels(PFE.final$AIC)))
## [1] -7620.414 -7618.398
MRK.final <- aiccall.list(MRK.armalist)
MRK.final
## AIC Order Code
## 1 -7573.916 c(5, 0, 5) 0
## 2 -7564.667 c(5, 0, 4) 0
## 3 -7566.636 c(5, 0, 3) 0
## 4 -7562.552 c(5, 0, 2) 0
## 5 -7564.404 c(5, 0, 1) 0
## 6 -7566.063 c(5, 0, 0) 0
## 7 -7564.566 c(4, 0, 5) 0
## 8 -7557.406 c(4, 0, 4) 0
## 9 -7559.397 c(4, 0, 3) 0
## 10 -7559.171 c(4, 0, 2) 0
## 11 -7566.566 c(3, 0, 5) 0
## 12 -7557.193 c(3, 0, 4) 0
## 13 -7559.193 c(3, 0, 3) 0
## 14 -7561.146 c(3, 0, 2) 0
## 15 -7560.342 c(2, 0, 5) 0
## 16 -7559.104 c(2, 0, 4) 0
## 17 -7561.158 c(2, 0, 3) 0
## 18 -7562.207 c(1, 0, 5) 0
## 19 -7564.009 c(0, 0, 5) 0
MRK.final[MRK.final$Code==1,] #no convergence issues
## [1] AIC Order Code
## <0 rows> (or 0-length row.names)
MRK.final[as.numeric(MRK.final$AIC)<=5,] #no clear AIC winners
## AIC Order Code
## 8 -7557.406 c(4, 0, 4) 0
## 10 -7559.171 c(4, 0, 2) 0
## 12 -7557.193 c(3, 0, 4) 0
## 13 -7559.193 c(3, 0, 3) 0
## 16 -7559.104 c(2, 0, 4) 0
range(as.numeric(levels(MRK.final$AIC)))
## [1] -7573.916 -7557.193
par(mfrow=c(2,1), oma=c(0,0,0,0))
tsdiag.new(PFE.arma10)
tsdiag.new(PFE.arma01)
paste(PFE.arma10$aic, PFE.arma01$aic)
## [1] "-7620.41436369809 -7620.36907928162"
# may as well choose AR1 -- lower AIC (not by much)
par(mfrow=c(2,2))
tsdiag.new(MRK.arma50)
tsdiag.new(MRK.arma32)
tsdiag.new(MRK.arma23)
tsdiag.new(MRK.arma05)
paste(MRK.arma50$aic, MRK.arma32$aic, MRK.arma23$aic, MRK.arma05$aic)
## [1] "-7566.06326450209 -7561.14593501032 -7561.15837102485 -7564.00913126507"
# AR5 is lowest -- again not by much