.First
.First <- function()
 invisible(library(Hmisc,T))
# .First()
page(describe(support),
   multi=T)
v <- varclus(~age + sex +
  dzgroup + num.co + scoma +
  race + meanbp + hrt + temp +
  pafi + alb, data=support)
plot(v)
s <- spearman2(totcst ~age + sex +
  dzgroup + num.co + scoma +
  race + meanbp + hrt + temp +
  pafi + alb, data=support)
s <- spearman2(totcst ~age + sex +
  dzgroup + num.co + scoma +
  race + meanbp + hrt + temp +
  pafi + alb, data=support,p=2)
s
plot(s)
n <- naclus(support)
plot(n)
naplot(n)
library(rpart)
f <- rpart(is.na(totchg) ~
 age + sex + dzgroup + num.co +
 scoma + race + meanbp,
 data=support)
f <- rpart(is.na(totcst) ~
 age + sex + dzgroup + num.co +
 scoma + race + meanbp,
 data=support)
plot(f)
text(f)
.First
.First <- function()
 invisible(library(Hmisc,T))
# .First()
page(describe(support),
   multi=T)
v <- varclus(~age + sex +
  dzgroup + num.co + scoma +
  race + meanbp + hrt + temp +
  pafi + alb, data=support)
plot(v)
s <- spearman2(totcst ~age + sex +
  dzgroup + num.co + scoma +
  race + meanbp + hrt + temp +
  pafi + alb, data=support)
s <- spearman2(totcst ~age + sex +
  dzgroup + num.co + scoma +
  race + meanbp + hrt + temp +
  pafi + alb, data=support,p=2)
s
plot(s)
n <- naclus(support)
plot(n)
naplot(n)
library(rpart)
f <- rpart(is.na(totchg) ~
 age + sex + dzgroup + num.co +
 scoma + race + meanbp,
 data=support)
f <- rpart(is.na(totcst) ~
 age + sex + dzgroup + num.co +
 scoma + race + meanbp,
 data=support)
plot(f)
text(f)
?varclus
x <- c(1,2,3,NA)
impute(x)
names(f)
f$cptable
table(race)
attach(support)
table(race)
describe(race)
race.i <- impute(race)
summary(race.i)
page(describe(support), multi=T)
w <- function(d) {
 page(names(d),multi=T)
 invisible() }
w(support)
fix(w)
w(support)
search()
xtrans <- transcan(~ totcst +
 age + sex + dzgroup +
 race + meanbp + hrt + temp +
 pafi + alb, n.impute=5)
levels(race)
names(attributes(xtrans))
page(attr(xtrans,'imputed')[1:10,])
page(attr(xtrans,'imputed'))
describe(totcst)
args(impute.transcan)
page(impute(xtrans,age,1))
page(impute(xtrans,totcst,1))
page(impute(xtrans,race,1))
page(impute(xtrans,race,2))
dd <- datadist(support)
library(Design,T)
dd <- datadist(support)
page(dd)
options(datadist='dd')
f <- ols(log(totcst) ~
 rcs(age,3) + sex + dzgroup +
 pol(num.co,2) + pol(scoma,2) +
 race + rcs(meanbp,4) +
 rcs(hrt,5) + rcs(temp,5) +
 rcs(pafi,3) + rcs(alb,3),
 subset=totcst > 0)
page(f)
page(f)
an <- anova(f)
page(an)
options(width=80)
page(an)
anova(f, hrt, temp, meanbp)
plot(an)
plot(f, temp=NA)
scat1d(temp)
plot(f, race=NA)
args(plot.Design)
plot(f, race=NA, method='dot')
plot(f, dzgroup=NA)
plot(f, dzgroup=NA, method='dot')
plot(f, temp=NA)
plot(f, temp=NA, fun=exp)
?smearingEst
smean <- function(yhat, res) {
 smearingEst(yhat, esp, res,
   statistic='mean') }
fix(smean)
smean$res <- resid(f)
describe(resid(f))
smean$res <- resid(f)[!is.na(
   resid(f))]
page(smean)
plot(f, temp=NA, fun=smean)
par(mfrow=c(3,4))
plot(f)
?predict.Design
g <- Function(f)
page(g)
?ols
page(g)
g()
g(sex='female')
g(age=50:60)
s <- summary(f)
page(s)
s <- summary(f, antilog=T)
page(s)
plot(s)
par(mfrow=c(1,1))
plot(s)
plot(s,log=T)
s <- summary(f, antilog=T,
   age=c(30,50,70))
plot(s,log=T)
nomogram(f, fun=list('Mean'=
 smean))
nomogram(f, fun=list('Mean'=
 smean,'Median'=exp))
f <- update(f,x=T,y=T)
dim(f$x)
length(f$y)
dimnames(f$x)[[2]]
v <- validate(f, B=40)
table(race)
race2 <- combine.levels(race)
table(race2)
f <- ols(log(totcst) ~
 rcs(age,3) + sex + dzgroup +
 pol(num.co,2) + pol(scoma,2) +
 race2 + rcs(meanbp,4) +
 rcs(hrt,5) + rcs(temp,5) +
 rcs(pafi,3) + rcs(alb,3),
 subset=totcst > 0,x=T,y=T
)
v <- validate(f, B=40)
set.seed(1);v <- validate(f, B=40, tol=1e-14)
set.seed(3)
v <- validate(f, B=40, tol=1e-14)
set.seed(5)
v <- validate(f, B=40, tol=1e-14)
set.seed(55)
v <- validate(f, B=40, tol=1e-14)
set.seed(555)
v <- validate(f, B=40, tol=1e-14)
set.seed(555)
v <- validate(f, B=37, tol=1e-14)
v
page(v)
set.seed(555)
cal <- calibrate(f, B=37, tol=1e-14)
plot(cal)
qqnorm(resid(f))
qqline(resid(f))
plot.lm(f)
xYplot(resid(f) ~ fitted(f),
 method='quantile')
xYplot(resid(f) ~ temp,
 method='quantile')
find(temp)
length(resid(f))
length(temp)
xYplot(resid(f) ~ temp[totcst>0],
 method='quantile')
bwplot(dzgroup[totcst>0] ~ resid(f))
bwplot(dzgroup[totcst>0] ~ 
  resid(f),panel=panel.bpplot)
args(panel.bpplot)
bwplot(dzgroup[totcst>0] ~ 
  resid(f),panel=panel.bpplot,
 nout=T)
?bpplot
?panel.bpplot
bwplot(dzgroup[totcst>0] ~ 
  resid(f),panel=panel.bpplot,
 nout=1000)
ecdf(~resid(f),groups=
 dzgroup[totcst>0])
ecdf(~resid(f),group=
 dzgroup[totcst>0])
page(f)
f$x <- f$y <- NULL
args(fit.mult.impute)
g <- fit.mult.impute(formula(f),
  ols, xtrans)
?fit.mult.impute
g <- fit.mult.impute(formula(f),
  ols, xtrans, subset=totcst>0)
totcst[totcst==0] <- 1
j <- attr(xtrans,'imputed')$totcst
dim(j)
sm(j==0)
sum(j==0)
j[j==0] <- 1
dim(j)
attr(xtrans,'imputed')$totcst <- 
 j
g <- fit.mult.impute(formula(f),
  ols, xtrans)
form_edit(formula(f))
form
g <- fit.mult.impute(formula(f),
  ols, xtrans)
g <- fit.mult.impute(form,
  ols, xtrans)
page(summary(xtrans))
form
g <- fit.mult.impute(form,
  ols, xtrans, data=support)
find(race)
args(fit.mult.impute)
fix(form)
g <- fit.mult.impute(form,
  ols, xtrans, data=support)
g <- fit.mult.impute(form,
  ols, xtrans)
names(support)
h <- lrm(hospdead ~ rcs(meanbp,
5)*rcs(hrt,5),x=T,y=T)
h <- lrm(hospdead ~ rcs(meanbp,
h <- lrm(hospdead ~ rcs(meanbp,
4)*rcs(hrt,4),x=T,y=T)
h <- lrm(hospdead ~ rcs(meanbp,
4)*rcs(hrt,4),x=T,y=T,tol=1e-14)
plot(h, meanbp=NA, hrt=NA,
method='image')
page(anova(h))
i <- pentrace(h, c(0,10,100,500))
i <- pentrace(h, c(0,10,100,500
),tol=1e-16)
i
?pentrace
ls()
rm(totcst)
rm(x,s,i,j,n,an)
f <- log(totcst+1) ~
 rcs(age,3) + sex + dzgroup +
 pol(num.co,2) + pol(scoma,2) +
 race + rcs(meanbp,4) +
 rcs(hrt,5) + rcs(temp,5) +
 rcs(pafi,3) + rcs(alb,3)
form <- f
g <- fit.mult.impute(form, ols,
 xtrans)
page(g)
f <- ols(log(totcst+1) ~
 rcs(age,3) + sex + dzgroup +
 pol(num.co,2) + pol(scoma,2) +
 race + rcs(meanbp,4) +
 rcs(hrt,5) + rcs(temp,5) +
 rcs(pafi,3) + rcs(alb,3))
describe(resid(f))
summary(is.na(resid) ~
 sex + age + dzgroup)
summary(is.na(resid(f)) ~
 sex + age + dzgroup)
ss_summary(is.na(resid(f)) ~
 sex + age + dzgroup)
page(ss)
vv <- tapply(resid(f), var,
 na.rm=T)
vm <- function(y)var(!is.na(y))
vv <- tapply(resid(f), vv)
vv <- tapply(resid(f), vm)
vv <- tapply(resid(f), dzgroup, vm)
plot(vv)
dotplot(names(vv) ~ vv)
vv
page(g)
page(anova(g))
plot(g, age=NA)
?impute.transcan
search()
detach(3)
dim(titanic3)
v <- c('pclass','survived',
 'age','sex','sibsp','parch')
page(describe(titanic3[,v]),
 multi=T)
dd <- datadist(titanic3[,v])
options(datadist='dd')
attach(titanic3[,v])
find(age)
find(sex)
plsmo(age, survived,
  group=interaction(pclass,sex),
  col=1:6, datadensity=T)
f <- lrm(survived ~
 (sex + pclass + rcs(age,5))^2+
 rcs(age,5)*sibsp)
f$stats
an <- anova(f)
plot(an)
plot(f, age=NA, pclass=NA,
 sex='male', conf.int=F)
plot(f, age=NA, pclass=NA,
 sex='female', conf.int=F)
plot(f, age=NA, pclass=NA,
 sex='female', conf.int=F,
 fun=plogis)
f <- update(f, x=T, y=T)
set.seed(131)
options(digits=2)
validate(f, B=80)
cal <- calibrate(f,B=25)
plot(cal)
na.patterns(titanic3)
na.patterns <- naclus(titanic3)
plot(na.patterns)
args(rpart)
who.na <- rpart(is.na(age) ~
 sex + pclass + survived +
 sibsp + parch, minbucket=15)
plot(who.na);text(who.na)
m <- lrm(is.na(age) ~
 sex*pclass + survived +
 sibsp + parch)
page(anova(m))
set.seed(17)
xtrans <- transcan(~ I(age) +
 sex + pclass + sibsp + parch +
 survived, n.impute=5, pl=F)
summary(xtrans)
attr(xtrans,'imputed)$age[1:10,]
'
'
attr(xtrans,'imputed')$age[1:10,]
formula(f)
f.mi <- fit.mult.impute(
 formula(f), lrm, xtrans)
coef(f.mi)
page(anova(f.mi))
pre.logit <- Function(f.mi)
page(pre.logit)
plogis(pre.logit(age=c(2,21,50),
sex='male',pclass='3rd'))
drep <- dataRep(~ roundN(age,10) +
 sex + pclass + roundN(sibsp,clip=0:1))
drep
roundN
args(page)
print(drep,long=T)
Dialog(fitPar('f.mi',lp=F,
 fun=list('Prob[Survival]'=
    plogis)),
   limits='data',
   basename='Titanic',
   vary=list(sex=levels(sex)),
    datarep=drep)
page(CallBack.Titanic)
runmenu.Titanic()
search()
detach(2)
attach(support[,Cs(age,sex,
dzgroup,num.co,scoma,race,
meanbp,hrt,temp,pafi,alb)])
search()[2]
s <- summary(hospdead~.)
detach(2)
attach(support[,Cs(age,sex,
dzgroup,num.co,scoma,race,
meanbp,hrt,temp,pafi,alb,hospdead)])
s <- summary(hospdead~.)
s <- summary(hospdead~age+sex+
 dzgroup+num.co+scoma+race+
 meanbp+hrt+temp+pafi+alb)
plot(s)
s <- summary(hospdead~age+sex+
 dzgroup+num.co+scoma+race)
s2 <- summary(hospdead ~
 meanbp+hrt+temp+pafi+alb)
plot(s)
plot(s2)
plsmo(age,hospdead,fun=qlogis)
plsmo(age,hospdead,fun=qlogis,
 group=dzclass)
plsmo(age,hospdead,fun=qlogis,
 group=dzgroup)
dd <- datadist(support)
sp <- spearman2(hospdead ~
age+sex+
sp <- spearman2(hospdead ~
age+sex+
 dzgroup+num.co+scoma+race+
 meanbp+hrt+temp+pafi+alb)
plot(sp)
sp <- spearman2(hospdead ~
age+sex+
 dzgroup+num.co+scoma+race+
 meanbp+hrt+temp+pafi+alb, p=2)
plot(sp)
race <- impute(race)
pafi <- impute(pafi, 333.3)
alb <- impute(alb, 3.5)
find(race)
table(is.impute(race))
table(is.imputed(race))
# support$race <- race
table(race)
race <- combine.levels(race,
 minlev=.1)
dd <- datadist(dd, race)
table(race)
.First
.First <- function()
 invisible(library(Hmisc,T))
# .First()
page(describe(support),
   multi=T)
v <- varclus(~age + sex +
  dzgroup + num.co + scoma +
  race + meanbp + hrt + temp +
  pafi + alb, data=support)
plot(v)
s <- spearman2(totcst ~age + sex +
  dzgroup + num.co + scoma +
  race + meanbp + hrt + temp +
  pafi + alb, data=support)
s <- spearman2(totcst ~age + sex +
  dzgroup + num.co + scoma +
  race + meanbp + hrt + temp +
  pafi + alb, data=support,p=2)
s
plot(s)
n <- naclus(support)
plot(n)
naplot(n)
library(rpart)
f <- rpart(is.na(totchg) ~
 age + sex + dzgroup + num.co +
 scoma + race + meanbp,
 data=support)
f <- rpart(is.na(totcst) ~
 age + sex + dzgroup + num.co +
 scoma + race + meanbp,
 data=support)
plot(f)
text(f)
?varclus
x <- c(1,2,3,NA)
impute(x)
names(f)
f$cptable
table(race)
attach(support)
table(race)
describe(race)
race.i <- impute(race)
summary(race.i)
page(describe(support), multi=T)
w <- function(d) {
 page(names(d),multi=T)
 invisible() }
w(support)
fix(w)
w(support)
search()
xtrans <- transcan(~ totcst +
 age + sex + dzgroup +
 race + meanbp + hrt + temp +
 pafi + alb, n.impute=5)
levels(race)
names(attributes(xtrans))
page(attr(xtrans,'imputed')[1:10,])
page(attr(xtrans,'imputed'))
describe(totcst)
args(impute.transcan)
page(impute(xtrans,age,1))
page(impute(xtrans,totcst,1))
page(impute(xtrans,race,1))
page(impute(xtrans,race,2))
dd <- datadist(support)
library(Design,T)
dd <- datadist(support)
page(dd)
options(datadist='dd')
f <- ols(log(totcst) ~
 rcs(age,3) + sex + dzgroup +
 pol(num.co,2) + pol(scoma,2) +
 race + rcs(meanbp,4) +
 rcs(hrt,5) + rcs(temp,5) +
 rcs(pafi,3) + rcs(alb,3),
 subset=totcst > 0)
page(f)
page(f)
an <- anova(f)
page(an)
options(width=80)
page(an)
anova(f, hrt, temp, meanbp)
plot(an)
plot(f, temp=NA)
scat1d(temp)
plot(f, race=NA)
args(plot.Design)
plot(f, race=NA, method='dot')
plot(f, dzgroup=NA)
plot(f, dzgroup=NA, method='dot')
plot(f, temp=NA)
plot(f, temp=NA, fun=exp)
?smearingEst
smean <- function(yhat, res) {
 smearingEst(yhat, esp, res,
   statistic='mean') }
fix(smean)
smean$res <- resid(f)
describe(resid(f))
smean$res <- resid(f)[!is.na(
   resid(f))]
page(smean)
plot(f, temp=NA, fun=smean)
par(mfrow=c(3,4))
plot(f)
?predict.Design
g <- Function(f)
page(g)
?ols
page(g)
g()
g(sex='female')
g(age=50:60)
s <- summary(f)
page(s)
s <- summary(f, antilog=T)
page(s)
plot(s)
par(mfrow=c(1,1))
plot(s)
plot(s,log=T)
s <- summary(f, antilog=T,
   age=c(30,50,70))
plot(s,log=T)
nomogram(f, fun=list('Mean'=
 smean))
nomogram(f, fun=list('Mean'=
 smean,'Median'=exp))
f <- update(f,x=T,y=T)
dim(f$x)
length(f$y)
dimnames(f$x)[[2]]
v <- validate(f, B=40)
table(race)
race2 <- combine.levels(race)
table(race2)
f <- ols(log(totcst) ~
 rcs(age,3) + sex + dzgroup +
 pol(num.co,2) + pol(scoma,2) +
 race2 + rcs(meanbp,4) +
 rcs(hrt,5) + rcs(temp,5) +
 rcs(pafi,3) + rcs(alb,3),
 subset=totcst > 0,x=T,y=T
)
v <- validate(f, B=40)
set.seed(1);v <- validate(f, B=40, tol=1e-14)
set.seed(3)
v <- validate(f, B=40, tol=1e-14)
set.seed(5)
v <- validate(f, B=40, tol=1e-14)
set.seed(55)
v <- validate(f, B=40, tol=1e-14)
set.seed(555)
v <- validate(f, B=40, tol=1e-14)
set.seed(555)
v <- validate(f, B=37, tol=1e-14)
v
page(v)
set.seed(555)
cal <- calibrate(f, B=37, tol=1e-14)
plot(cal)
qqnorm(resid(f))
qqline(resid(f))
plot.lm(f)
xYplot(resid(f) ~ fitted(f),
 method='quantile')
xYplot(resid(f) ~ temp,
 method='quantile')
find(temp)
length(resid(f))
length(temp)
xYplot(resid(f) ~ temp[totcst>0],
 method='quantile')
bwplot(dzgroup[totcst>0] ~ resid(f))
bwplot(dzgroup[totcst>0] ~ 
  resid(f),panel=panel.bpplot)
args(panel.bpplot)
bwplot(dzgroup[totcst>0] ~ 
  resid(f),panel=panel.bpplot,
 nout=T)
?bpplot
?panel.bpplot
bwplot(dzgroup[totcst>0] ~ 
  resid(f),panel=panel.bpplot,
 nout=1000)
ecdf(~resid(f),groups=
 dzgroup[totcst>0])
ecdf(~resid(f),group=
 dzgroup[totcst>0])
page(f)
f$x <- f$y <- NULL
args(fit.mult.impute)
g <- fit.mult.impute(formula(f),
  ols, xtrans)
?fit.mult.impute
g <- fit.mult.impute(formula(f),
  ols, xtrans, subset=totcst>0)
totcst[totcst==0] <- 1
j <- attr(xtrans,'imputed')$totcst
dim(j)
sm(j==0)
sum(j==0)
j[j==0] <- 1
dim(j)
attr(xtrans,'imputed')$totcst <- 
 j
g <- fit.mult.impute(formula(f),
  ols, xtrans)
form_edit(formula(f))
form
g <- fit.mult.impute(formula(f),
  ols, xtrans)
g <- fit.mult.impute(form,
  ols, xtrans)
page(summary(xtrans))
form
g <- fit.mult.impute(form,
  ols, xtrans, data=support)
find(race)
args(fit.mult.impute)
fix(form)
g <- fit.mult.impute(form,
  ols, xtrans, data=support)
g <- fit.mult.impute(form,
  ols, xtrans)
names(support)
h <- lrm(hospdead ~ rcs(meanbp,
5)*rcs(hrt,5),x=T,y=T)
h <- lrm(hospdead ~ rcs(meanbp,
h <- lrm(hospdead ~ rcs(meanbp,
4)*rcs(hrt,4),x=T,y=T)
h <- lrm(hospdead ~ rcs(meanbp,
4)*rcs(hrt,4),x=T,y=T,tol=1e-14)
plot(h, meanbp=NA, hrt=NA,
method='image')
page(anova(h))
i <- pentrace(h, c(0,10,100,500))
i <- pentrace(h, c(0,10,100,500
),tol=1e-16)
i
?pentrace
ls()
rm(totcst)
rm(x,s,i,j,n,an)
f <- log(totcst+1) ~
 rcs(age,3) + sex + dzgroup +
 pol(num.co,2) + pol(scoma,2) +
 race + rcs(meanbp,4) +
 rcs(hrt,5) + rcs(temp,5) +
 rcs(pafi,3) + rcs(alb,3)
form <- f
g <- fit.mult.impute(form, ols,
 xtrans)
page(g)
f <- ols(log(totcst+1) ~
 rcs(age,3) + sex + dzgroup +
 pol(num.co,2) + pol(scoma,2) +
 race + rcs(meanbp,4) +
 rcs(hrt,5) + rcs(temp,5) +
 rcs(pafi,3) + rcs(alb,3))
describe(resid(f))
summary(is.na(resid) ~
 sex + age + dzgroup)
summary(is.na(resid(f)) ~
 sex + age + dzgroup)
ss_summary(is.na(resid(f)) ~
 sex + age + dzgroup)
page(ss)
vv <- tapply(resid(f), var,
 na.rm=T)
vm <- function(y)var(!is.na(y))
vv <- tapply(resid(f), vv)
vv <- tapply(resid(f), vm)
vv <- tapply(resid(f), dzgroup, vm)
plot(vv)
dotplot(names(vv) ~ vv)
vv
page(g)
page(anova(g))
plot(g, age=NA)
?impute.transcan
search()
detach(3)
dim(titanic3)
v <- c('pclass','survived',
 'age','sex','sibsp','parch')
page(describe(titanic3[,v]),
 multi=T)
dd <- datadist(titanic3[,v])
options(datadist='dd')
attach(titanic3[,v])
find(age)
find(sex)
plsmo(age, survived,
  group=interaction(pclass,sex),
  col=1:6, datadensity=T)
f <- lrm(survived ~
 (sex + pclass + rcs(age,5))^2+
 rcs(age,5)*sibsp)
f$stats
an <- anova(f)
plot(an)
plot(f, age=NA, pclass=NA,
 sex='male', conf.int=F)
plot(f, age=NA, pclass=NA,
 sex='female', conf.int=F)
plot(f, age=NA, pclass=NA,
 sex='female', conf.int=F,
 fun=plogis)
f <- update(f, x=T, y=T)
set.seed(131)
options(digits=2)
validate(f, B=80)
cal <- calibrate(f,B=25)
plot(cal)
na.patterns(titanic3)
na.patterns <- naclus(titanic3)
plot(na.patterns)
args(rpart)
who.na <- rpart(is.na(age) ~
 sex + pclass + survived +
 sibsp + parch, minbucket=15)
plot(who.na);text(who.na)
m <- lrm(is.na(age) ~
 sex*pclass + survived +
 sibsp + parch)
page(anova(m))
set.seed(17)
xtrans <- transcan(~ I(age) +
 sex + pclass + sibsp + parch +
 survived, n.impute=5, pl=F)
summary(xtrans)
attr(xtrans,'imputed)$age[1:10,]
'
'
attr(xtrans,'imputed')$age[1:10,]
formula(f)
f.mi <- fit.mult.impute(
 formula(f), lrm, xtrans)
coef(f.mi)
page(anova(f.mi))
pre.logit <- Function(f.mi)
page(pre.logit)
plogis(pre.logit(age=c(2,21,50),
sex='male',pclass='3rd'))
drep <- dataRep(~ roundN(age,10) +
 sex + pclass + roundN(sibsp,clip=0:1))
drep
roundN
args(page)
print(drep,long=T)
Dialog(fitPar('f.mi',lp=F,
 fun=list('Prob[Survival]'=
    plogis)),
   limits='data',
   basename='Titanic',
   vary=list(sex=levels(sex)),
    datarep=drep)
page(CallBack.Titanic)
runmenu.Titanic()
search()
detach(2)
attach(support[,Cs(age,sex,
dzgroup,num.co,scoma,race,
meanbp,hrt,temp,pafi,alb)])
search()[2]
s <- summary(hospdead~.)
detach(2)
attach(support[,Cs(age,sex,
dzgroup,num.co,scoma,race,
meanbp,hrt,temp,pafi,alb,hospdead)])
s <- summary(hospdead~.)
s <- summary(hospdead~age+sex+
 dzgroup+num.co+scoma+race+
 meanbp+hrt+temp+pafi+alb)
plot(s)
s <- summary(hospdead~age+sex+
 dzgroup+num.co+scoma+race)
s2 <- summary(hospdead ~
 meanbp+hrt+temp+pafi+alb)
plot(s)
plot(s2)
plsmo(age,hospdead,fun=qlogis)
plsmo(age,hospdead,fun=qlogis,
 group=dzclass)
plsmo(age,hospdead,fun=qlogis,
 group=dzgroup)
dd <- datadist(support)
sp <- spearman2(hospdead ~
age+sex+
sp <- spearman2(hospdead ~
age+sex+
 dzgroup+num.co+scoma+race+
 meanbp+hrt+temp+pafi+alb)
plot(sp)
sp <- spearman2(hospdead ~
age+sex+
 dzgroup+num.co+scoma+race+
 meanbp+hrt+temp+pafi+alb, p=2)
plot(sp)
race <- impute(race)
pafi <- impute(pafi, 333.3)
alb <- impute(alb, 3.5)
find(race)
table(is.impute(race))
table(is.imputed(race))
# support$race <- race
table(race)
race <- combine.levels(race,
 minlev=.1)
dd <- datadist(dd, race)
table(race)
search()[1:3]
detach(2)
v
attach(support)
s <- spearman2(totcst ~
 dzgroup + age + scoma)
s
options(digits=4)
s
82.1*7
lrm(cut2(totcst,g=2) ~ dzgroup)
lrm(cut2(totcst,g=3) ~ dzgroup)
lrm(cut2(totcst,g=4) ~ dzgroup)$stats
lrm(cut2(totcst,g=5) ~ dzgroup)$stats
lrm(cut2(totcst,g=7) ~ dzgroup)$stats
lrm(cut2(totcst,g=10) ~ dzgroup)$stats
lrm(cut2(totcst,g=14) ~ dzgroup)$stats
lrm(cut2(totcst,g=20) ~ dzgroup)$stats
page(lrm(cut2(totcst,g=20) ~ dzgroup))
f <- lrm(cut2(totcst,g=20) ~ dzgroup +
   rcs(age,3) + rcs(meanbp,5) +
   scoma)
page(anova(f))
dd <- datadist(support)
options(datadist='dd')
plot(f)
specs(f)
abline(v=c(47,128))
fd <- lrm(cut2(totcst,g=20) ~ dzgroup)
plot(f,dzgroup=NA)
plot(fd,dzgroup=NA)
plot(f,dzgroup=NA)
plot(fd,dzgroup=NA,add=T,col=3)
?survplot
?event.chart
describe(d.time)
d.time <- d.time/365.25
unis(d.time) <- 'Year'
units(d.time) <- 'Year'
S <- Surv(d.time,death)
S[1:20,]
survplot(survfit(S),n.risk=T)
survplot(survfit(S),fun=qnorm,
 logt=T)
survplot(survfit(S)~dzclass,fun=qnorm,
 logt=T)
survplot(survfit(S~dzclass),fun=qnorm,
 logt=T)
sub <- dzclass!='cancer'
survplot(survfit(S~dzclass,
 data=sub),fun=qnorm,logt=T)
survplot(survfit(S~dzclass,
 subset=sub),fun=qnorm,logt=T)
search()
detach(2)
acute <- support$dzclass %in%
 c('ARF/MOSF','Coma')
table(acute)
attach(support[acute,])
remove(c('d.time','S'))
d.time <- d.time/365.25
units(d.time) <- 'Year'
S <- Surv(d.time,death)
dim(S)
survplot(survfit(S ~ dzgroup),
 fun=qnorm, logt=T)
f <- psm(S ~ dzgroup +
 rcs(age,5) + rcs(meanbp,5),
 dist='gau', y=T)
dd <- datadist(age,num.co,
 scoma, meanbp, hrt, resp, temp,
 crea, sod, adlsc, wblc, pafi,ph,
dzgroup,race)
find(pafi)
remove('pafi','race')
dd <- datadist(age,num.co,
 scoma, meanbp, hrt, resp, temp,
 crea, sod, adlsc, wblc, pafi,ph,
dzgroup,race)
options(datadist='dd')
f <- psm(S ~ dzgroup +
 rcs(age,5)+rcs(meanbp,5),
 dist='ga', y=T)
page(f)
latex(f, file='/windows/temp/f.tex')
latex(f, fi='/windows/temp/f.tex')
win3('notepad /windows/temp/f.tex')
page(anova(f))
r <- resid(f)
dim(r)
class(r)
r[1:10,]
plot(f, meanbp=T)
plot(f, meanbp=NA)
survplot(survfit(r))
survplot(r)
survplot(r, dzgroup)
survplot(r, age)
survplot(r, meanbp)
random.number <- runif(length(
 age))
survplot(r, random.number)
mean(death)
sum(death)
min(d.time[death==0])
d.timet <- pmin(d.time,.9747)
w <- spearman2(d.timet ~ age +
 num.co + scoma+ meanbp +
 hrt + resp + temp + crea + 
 sod + adlsc + wblc + pafi + ph +
 dzgroup + race, p=2)
plot(w)
wbc <- impute(wbc,9)
wblc <- impute(wblc,9)
pafi <- impute(pafi, 333.3)
ph <- impute(ph, 7.4)
race <- combine.levels(race)
table(race)
white <- race=='white'
dd <- datadist(dd, white)
f <- psm(S ~ age+sex+dzgroup+
 pol(num.co,2)+pol(scoma,2)+
 pol(adlsc,2)+white+rcs(meanbp,5)+
 rcs(hrt,3)+resp+rcs(temp,3)+
 rcs(crea,5)+sod+rcs(wblc,3)+
 rcs(pafi,3), dist='gau')
table(white)
args(psm)
?psm
?survreg
f <- psm(S ~ rcs(age,3)+sex+dzgroup+
 pol(num.co,2)+pol(scoma,2)+
 pol(adlsc,2)+white+rcs(meanbp,5)+
 rcs(hrt,3)+resp+rcs(temp,3)+
 rcs(crea,5)+sod+rcs(wblc,3)+
 rcs(pafi,3), dist='gau')
page(f)
page(anova(f))
par(mfrow=c(3,5))
plot(f,ref.zero=T,ylim=c(-3,2))
page(dd)
?datadist
ddold <- dd
dd <- datadist(f)
plot(f,ref.zero=T,ylim=c(-3,2))
par(mfrow=c(1,1));plot(f,crea=NA)
scat1d(crea)
page(anova(f))
plot(anova(f))
s
sp
sc
plot(w)
s <- summary(f)
plot(s,log=T)
rcorr.cens(age,S)
rcorr.cens(meanbp,S)
rcorr.cens(fitted(f),S)
g <- update(f,x=T,y=T)
validate(g,B=20)
validate(g,B=20)
set.seed(1);validate(g,B=10)
set.seed(1);validate(g,B=10,dxy=T)
cal <- calibrate(f, u=1, m=60)
cal <- calibrate(g, u=1, m=60)
cal <- calibrate(g, u=1, m=60, B=1-)
cal <- calibrate(g, u=1, m=60, B=10)
plot(cal)
?val.surv
Z <- predict(f)
form <- formula(f)
fix(form)
a <- ols(form, sigma=1)
page(a)
fbw <- fastbw(a, aics=10000)
page(fwb)
page(fbw)
expected.surv <- Mean(f)
page(expected.surv)
quantile.surv <- Quantile(f)
page(quantile.surv)
f <- areg.boot(totcst ~
 age + meanbp + dzgroup +
 white + scoma, B=50)
plot(f)
page(f)
ftrans <- Function(f)
names(ftrans)
length(ftrans)
page(ftrans$totcst)
plot(totcst, ftrans$totcst(totcst))
plot(ftrans$totcst(totcst),
 log(totcst))
qqnorm(resid(f))
qqline(resid(f))
xYplot(resid(f) ~ fitted(f),
 method='quantile')
xYplot(resid(f) ~ fitted(f),
 method='quantile', m=100)
?xYplot
xYplot(resid(f) ~ fitted(f),
 method='quantile', nx=80)
xYplot(resid(f) ~ fitted(f),
 method='quantile', nx=80,
 xlim=c(10000,60000))
args(page)
page
?predict.areg.boot
formula(f)
names(f)
f$call
predict(f,data.frame(age=30,
 meanbp=50, dzgroup='ARF/MOSF',
 white=T,scoma=0),
 statistic='mean')
predict(f,data.frame(age=30,
 meanbp=50, dzgroup='ARF/MOSF',
 white=T,scoma=0),
 statistic='median')
s <- summary(f)
class(f)
?areg.boot
summary(f)
traceback()
s <- summary(f,statistic='median')
s <- summary(f,statistic='mean')
?areg.boot
f <- areg.boot(totcst ~
 monotone(age) + meanbp + dzgroup +
 white + scoma, B=40
)
plot(f)