cs-5821/hw3/Rhistory
2020-12-23 13:24:59 -05:00

463 lines
12 KiB
Plaintext

auto = read.table("auto.data",header=T,na.strings="?")
length(x=auto$mpg)
glm
glm.pred
help(rep)
glm.pred=rep(FALSE,397)
glm.pred
medium(auto$mpg)
median(auto$mpg)
glm.pred[auto$mpg>median(auto$mpg)]=T
glm.pred
contour(auto)
contour(glm.pred ~ auto$mpg)
contour(glm.pred,auto$mpg)
help(contour)
contour(auto$mpg,auto$horsepower,glm.pred)
glm.pred
length(glm.pred)
table(glm.pred,auto$mpg)
table(glm.pred,auto$mpg,auto$horsepower)
glm.pred=rep(0,397)
glm.pred[auto$mpg>median(auto$mpg)]=1
glm.pred
auto$mpg01=rep(0,397)
auto$mpg01[auto$mpg>median(auto$mpg)]=1
auto$mpg01
auto$mpg01
auto$mpg01
plots(auto)
plot(auto)
boxplot(auto)
boxplot.matrix(auto)
help(boxplot)
boxplot(auto$mpg01,auto)
boxplot(auto$mpg,auto)
boxplot(auto$mpg)
boxplot(auto)
boxplot(mpg01 ~ auto)
boxplot(mpg01 ~)
boxplot(auto$mpg01 ~ auto)
attach(auto)
boxplot(mpg01)
boxplot(mpg01 ~ auto)
boxplot(mpg01 ~ auto,auto)
boxplot(mpg01 ~ auto,data = auto)
help(plot.table)
plot.table(auto)
help(plot.table)
plot(auto)
plot(auto,t="box")
help(plot.table)
help(plot.table,plot.frame=1)
help(plot.table)
help(plot.table,frame.plot=1)
help(plot.table)
help(plot.table,frame.plot=is.num)
help(plot.table)
plot(auto,t="box",frame.plot=1)
plot(auto,frame.plot=1)
plot(auto,frame.plot=1)
plot(auto,frame.plot=is.num)
plot(auto,frame.plot=0)
plot(auto,frame.plot="0")
plot(auto,frame.plot="1")
plot(auto,frame.plot=TRUE)
plot(auto,frame.plot=FALSE)
plot(auto,frame.plot=TRUE)
plot(auto,frame.plot=T)
plot(auto,frame.plot=1)
boxplot(mpg~mpg01,auto)
boxplot(mpg01 ~ mpg,auto)
boxplot(mpg01 ~ *,auto)
boxplot(mpg01 ~ ,auto)
boxplot(mpg01 ~ auto,auto)
boxplot(mpg01,auto)
boxplot(auto)
boxplot(auto,y=mpg01)
boxplot(auto,y=mpg)
boxplot(data = auto)
boxplot(auto)
help(for)
plot(auto,frame.plot=1)
plot(auto)
names(auto)
auto$name
help(sample)
x <- 1:12
x
sample(x)
help(sample)
sample(x,replace=T)
sample(x,replace=T)
sample(x,replace=F)
c
x
sample(x,replace=T)
x
help(sample)
sample(x[x>9])
sample(x[x>8])
help(sample)
x <- 1:10
sample(x[x>8])
sample(x[x>])
help(sample)
help(sample)
help(sample)
sample(auto,size=length(mpg01)/2)
x <- length(mpg01)
sample(x,size=length(mpg01)/2)
auto[sample(x,size=length(mpg01)/2)]
auto$mpg[sample(x,size=length(mpg01)/2)]
help(data.frame)
data.frame(
help(data.frame)
auto[sample(x,size=length(mpg01)/2)]
train = sample(x,size=length(mpg01)/2)
train =
auto[train]
auto$mpg[train]
auto$mpg[train,]
auto$mpg[train]
auto$mpg[23]
auto$mpg[228]
auto$mpg[391]
auto.test=auto[!train]
auto.train=auto[train]
auto.test
summary(auto.test)
train=(mpg<15)
train
train = (sample(x,size=length(mpg01)/2))
train
head(auto)
auto[,train[
auto[,train]
train
help(contains)
auto[1,train]
train
auto[[,train]]
auto[[1,train]]
autoi
head(auto)
head(auto[sample(nrow(auto),397/2)])
head(auto[sample(nrow(auto),3)])
data = data.frame(auto)
data
head(data[sample(nrow(data),3)])
nrow(data)
head(data[sample(ncol(data),3)])
head(data[sample(ncol(data),397/2)])
head(data[sample(ncol(data),3)])
head(data[sample(ncol(data),3)])
head(data[sample(ncol(data),3)])
head(data[sample(ncol(data),3)])
head(data[,sample(ncol(data),3)])
head(data[,sample(ncol(data),3)])
head(data[,sample(ncol(data),3)])
head(data[,sample(ncol(data),3)])
head(data[,sample(ncol(data),3)])
head(data[sample(ncol(data),3),])
head(data[sample(ncol(data),3),])
head(data[sample(ncol(data),3),])
head(data[sample(nrow(data),3),])
head(data[sample(nrow(data),397/2),])
head(data[sample(nrow(data),397/2),])
head(data[sample(nrow(data),397/2),])
head(data[sample(nrow(data),397/2),])
head(data[sample(nrow(data),397/2),])
head(auto[sample(nrow(auto),397/2),])
head(auto[sample(nrow(auto),397/2),])
head(auto[sample(nrow(auto),397/2),])
head(auto[sample(nrow(auto),397/2),])
head(auto[sample(nrow(auto),397/2),])
head(auto[sample(nrow(auto),397/2),])
head(auto[sample(nrow(auto),397/2),])
train = auto[sample(nrow(auto),397/2),]
[sample(nrow(auto),397/2),]
sample(nrow(auto),397/2)
train sample(nrow(auto),397/2)
train = sample(nrow(auto),397/2)
autp[train,]
auto[train,]
train = sample(nrow(auto),397/2)
head(auto[train,])
head(auto[!train,])
traindata = auto[train,]
testdata = auto[!train,]
testdata
traindata
length(traindata)
length(traindata$mpg)
198*2
summary(testdata)
testdata = auto[!train]
testdata
testdata = auto[!train,]
train
summary(train)
names(train)
head(traindata)
testdata = auto[!train,]
testdata
!train
train
?sample
sort(train)
train_vals = train
train = rep(false,397)
train = rep(F,397)
train
help for
?for
?for
help)for)
help(for)
help(for)
help lapply()
?lapply
sapply(train,
?sapply
sapply(train,
?sapply
train[train_vals]=T
train
traindata = auto[train,]
traindata
length(auto)
length(traindata)
length(traindata$mpg)
testdata=auto[!train,]
length(testdate$mpg)
length(testdata$mpg)
training_indices = sample(nrow(auto),397/2)
train_bools = rep(F,length(auto$mpg))
train_bools[training_indices]=T
head(train_bools)
length(train_bools)
train_data = auto[train_bools,]
test_data = auto[!train_bools,]
summary(train_data)
summary(test_data)
lda.fit
library(MASS)
lda.fit
lda()
detach(auto)
mpg01
mpg
attach(test_data)
mpg01
names()
names(test_data)
ldf.fit=lda(mpg01 ~ horsepower + weight + acceleration + displacement,data=test_data)
detach(test_data)
ldf.fit=lda(mpg01 ~ horsepower + weight + acceleration + displacement,data=test_data)
lda.fit
lda.fit=lda(mpg01 ~ horsepower + weight + acceleration + displacement,data=test_data)
lda.fit
summary(lda.fit)
coefficients(lda.fit)
plot(lda.fit)
lda.pred=predict(lda.fit,test_data)
lda.pred=predict(lda.fit, !training_bools)
lda.pred=predict(lda.fit, !training_indices)
test_data
lda.pred=predict(lda.fit, test_data)
lda.pred
plot(lda.pred)
names(lda.pred)
lda.class=lda.pres$class
lda.class=lda.pred$class
table(lda.class,testdata)
table(lda.class,test_data)
length(lda.class)
length(test_data)
table(lda.class,test_data$mpg01)
mean(lda.class==test_data$mpg01)
sum(lda.pred$posterior[,1]>=.5)
sum(lda.pred$posterior[,1]<.5)
lda.pred$posterior[,1]
sum(lda.pred$posterior<.5)
lda.pred$posterior
lda.pred$posterior<5
lda.pred$posterior<.5
sum(lda.pred$posterior<.5)
sum(lda.pred$posterior<.5[,1])
sum(lda.pred$posterior<.5[1])
sum(lda.pred$posterior<.5[2])
lda.pred$posterior<.5[2]
lda.pred$posterior<.5
lda.pred$posterior
lda.pred$posterior[,1]
lda.pred$posterior[1,]
lda.pred$posterior[,2]
lda.pred$posterior[,1]
lda.pred$posterior[,1]>.5
sum(lda.pred$posterior[,1]>.5)
sum.bool(lda.pred$posterior[,1]>.5)
?sum
sum.bool(lda.pred$posterior[,1]>.5,na.rm=T)
sum(lda.pred$posterior[,1]>.5,na.rm=T)
sum(lda.pred$posterior[,1]>.5)
sum(lda.pred$posterior[,1]>.5,na.rm=T)
sum(lda.pred$posterior[,1]>=.5,na.rm=T)
sum(lda.pred$posterior[,1]<.5,na.rm=T)
mean(lda.pred$[,1]==test_data,na.rm=T)
lda.pred
lda.pred$class
lda.pred$class==test_data$mpg01
mean(lda.pred$class==test_data$mpg01,na.rm=T)
mean(lda.pred$class!=test_data$mpg01,na.rm=T)
lda.fit=lda(mpg01 ~ horsepower + weight + acceleration + displacement,data=train_data)
lda.fit
mean(lda.pred$class==test_data$mpg01,na.rm=T)
lda.pred=predict(lda.fit, test_data)
mean(lda.pred$class==test_data$mpg01,na.rm=T)
mean(lda.pred$class!=test_data$mpg01,na.rm=T)
train_data == test_data
train_data$mpg01 == test_data$mpg01
lda.fit=lda(mpg01 ~ horsepower + weight + acceleration + displacement,data=train_data)
lda.pred=predict(lda.fit, test_data)
mean(lda.pred$class!=test_data$mpg01,na.rm=T)
lda.pred
lda.pred$posterior[,1]
summary(lda.fit)
lda.fit
lda.fit=lda(mpg01 ~ horsepower + weight + acceleration + displacement,data=test_data)
lda.fit
mean(lda.pred$class!=test_data$mpg01,na.rm=T)
lda.pred=predict(lda.fit, test_data)
mean(lda.pred$class!=test_data$mpg01,na.rm=T)
head(lda.pred)
lda.fit=lda(mpg01 ~ horsepower + weight + acceleration + displacement,data=train_data)
lda.pred=predict(lda.fit, test_data)
head(lda.pred)
mean(lda.pred$class!=test_data$mpg01,na.rm=T)
qda.fit=qda(mpg01 ~ horsepower + weight + acceleration + displacement,data=train_data)
qda.fit
qda.class=predict(qda.fit,test_data)$class
qda.class=predict(qda.fit,test_data,na.rm=T)$class
qda.class=predict(qda.fit,test_data)$class
qda.class
mean(qda.pred$class!=test_data$mpg01,na.rm=T)
qda.pred=predict(qda.fit,test_data)
qda.pred=predict(qda.fit,test_data,na.rm=T)
mean(qda.pred$class!=test_data$mpg01,na.rm=T)
glm.fit=glm(mpg01 ~ horsepower + weight + acceleration + displacement,data=train_data,family=binomial)
glm.probs=predict(glm.fit,test_data,type="response")
glm.pred=rep(0,199)
glm.pred[glm.probs>.5]=1
table(glm.pred,test_data$mpg01)
mean(glm.pred!=test_data$mpg01)
library(class)
?cbind
?knn
knn.fit = knn(train_data,test_data,auto$mpg01[training_indices])
knn.fit = knn(train_data,test_data,auto$mpg01[training_indices],k=1)
knn.fit = knn(train_data,test_data,auto$mpg01[training_indices],k=1)
?knn
training_indices
train_bools
knn.fit = knn(train_data,test_data,auto$mpg01[train_bools],k=1)
sdf = (mpg01<1)
sdf = (auto$mpg01<1)
sdf
train_bools
cbind(horsepower,displacement)
cbind(train_data$horsepower,displacement)
cbind(train_data$horsepower,train_data$displacement)
cbind(auto$horsepower,auto$displacement)[train_bools]
cbind(auto$horsepower,auto$displacement)[train_bools,]
cbind(auto$horsepower,auto$displacement,auto$weight,auto$acceleration)[train_bools,]
cbind(auto$horsepower,auto$displacement,auto$weight,auto$acceleration)[train_bools,]
train.X = cbind(auto$horsepower,auto$displacement,auto$weight,auto$acceleration)[train_bools,]
test.X = cbind(auto$horsepower,auto$displacement,auto$weight,auto$acceleration)[!train_bools,]
train.X = cbind(auto$horsepower,auto$displacement,auto$weight,auto$acceleration)[train_bools,]
test.X = cbind(auto$horsepower,auto$displacement,auto$weight,auto$acceleration)[!train_bools,]
train.mpg01 = auto[train_bools]
train.mpg01 = auto$mpg01[train_bools]
test.X = cbind(auto$horsepower,auto$displacement,auto$weight,auto$acceleration)[!train_bools,]
train.X = cbind(auto$horsepower,auto$displacement,auto$weight,auto$acceleration)[train_bools,]
test.X = cbind(auto$horsepower,auto$displacement,auto$weight,auto$acceleration)[!train_bools,]
train.mpg01 = auto$mpg01[train_bools]
set.seed(56)
knn.pred = knn(train.X,test.X,train.mpg01,k=1)
?cbind
?Knn
?knn
train.X = cbind(auto$horsepower,auto$displacement,auto$weight,auto$acceleration)[train_bools,]
test.X = cbind(auto$horsepower,auto$displacement,auto$weight,auto$acceleration)[!train_bools,]
train.mpg01 = auto$mpg01[train_bools]
train.X = train.X[!is.na(train.X)]
test.X = data.frame(test.X,
train.mpg01 = train.mpg01[!is.na(train.mpg01)]
knn.pred = knn(train.X,test.X,train.mpg01,k=1)
length(train.mpg01)
length(test.X)
text.X
test.X
test.X = cbind(auto$horsepower,auto$displacement,auto$weight,auto$acceleration)[!train_bools,]
length(test.X)
test.X
knn.pred = knn(train.X,test.X,train.mpg01,k=1)
train.X
train.X = cbind(auto$horsepower,auto$displacement,auto$weight,auto$acceleration)[train_bools,]
train.X
test.X
knn.pred = knn(train.X,test.X,train.mpg01,k=1)
?knn
length(train.X)
length(train.X[1,])
length(train.X[,1])
?knn
plot(auto)
train.X = cbind(auto$horsepower,auto$displacement)[train_bools,]
test.X = cbind(auto$horsepower,auto$displacement)[!train_bools,]
train.mpg01 = auto$mpg01[train_bools]
knn.pred = knn(train.X,test.X,train.mpg01,k=1)
train.X
test.X
train.mpg01
knn.pred = knn(train.X,test.X,train.mpg01,k=1)
q()
train
train.X
train.X
test.X
p1 = seq(1:1))
p1 = 1:10
p1
p1 = ,1:10
p2 = 10:20
p2
cbind(p1,p2)
p3=c(1,2,3,4,5,7,9,8,10)
p4=c(10,11,12,13,14,15,16,17,18,29,20)
px = cbind(p1,p2)
py = cbind(p3,p4)
py
?formula
test.X = cbind(auto$horsepower,auto$displacement,auto$weight,auto$acceleration)[!train_bools,]
train.X = cbind(auto$horsepower,auto$displacement,auto$weight,auto$acceleration)[train_bools,]
train.mpg01 = auto$mpg01[train_bools]
knn.pred = knn(train.X,test.X,train.mpg01,k=1)
import(library)
library(MASS)
knn.pred = knn(train.X,test.X,train.mpg01,k=1)
library(library)
library(class)
knn.pred = knn(train.X,test.X,train.mpg01,k=1)
knn.pred
summary(test_data)
summary(train_data)
fix(train_data)
?fix
fix(test_data)
test_data.fix
test_data.fix()
fix(test_data)
q()