#1
library(quantmod)
getSymbols('KL', source='yahoo', from='2010-01-01')
head(KL)
rt=diff(log(as.numeric(KL$KL.Adjusted)))
rt[is.na(rt)] = 0
rt[rt>0] <- 1
rt[rt<=0] <- 0
head(rt)

getSymbols('^GSPC', source='yahoo', from='2010-01-01')
mt=diff(log(as.numeric(GSPC$GSPC.Adjusted)))
head(mt)
mt[is.na(mt)] = 0
mt[mt>0] <- 1
mt[mt<=0] <- 0
head(mt)

#1.a
train=rt[1:1500]
frc=rt[1501:1963]
train0=train[4:1500]
train1=train[3:1499]
train2=train[2:1498]
train3=train[1:1497]

mt1=mt[3:1499]
mt2=mt[2:1498]
mt3=mt[1:1497]

df=data.frame(train0, train1, train2, train3, mt1,mt2,mt3)
dfrc=data.frame(frc)

#1.b
model=lm(train0 ~., data=df )
summary(model)

#1.c
library(forecast)
forecast.lm(object = model, newdata = dfrc)
plot(predict.lm(object = model))

#1.d
library(MASS)
model1=lda(train0~., data=df)

#1.e
model2=qda(train0~., data=df)
model2