# Q2 example
# visualization
interaction.plot(dat.Q2$med,dat.Q2$ingred,dat.Q2$bp,type='o',
   col = c("skyblue", "coral"),lwd =3, cex = 2,
   pch=c(18,19),xlab = "quantitily", ylab = "Blood Pressure",
   trace.label="type of medicine",legend = T)
# descriptive stats
tapply(dat.Q2$bp,list(dat.Q2$med,dat.Q2$ingred),mean)
# ANOVA
dat.Q2.aov = aov(bp~med*ingred,data=dat.Q2)
summary(dat.Q2.aov)
# Linear Regression
dat.Q2.lm = lm(bp~med*ingred, contrasts=list(med=contr.poly), data=dat.Q2)
summary(dat.Q2.lm)
anova(dat.Q2.lm)

# Q3 example 1
# visualization
plot(hours~btype, data= dat.Q3V1)
# ANOVA
dat.Q3V1.aov = aov(hours~btype,data =dat.Q3V1)
summary(dat.Q3V1.aov)
TukeyHSD(dat.Q3V1.aov)
# Linear Regression
dat.Q3V1.lm = lm(hours~btype, data = dat.Q3V1)
summary(dat.Q3V1.aov)

# Q3 example 2
# visualization
plot(dat.Q3V1$hours, dat.Q3V1$A,type = 'p',pch=20,cex =2, 
   xlab = "hours cleaned", ylab= "Blood Type",yaxt ="n")
axis(2, at = c(0,1), labels = c("not A", "A"))
# GLM
dat.Q3V1.glm=glm(A~hours,family="binomial",data=dat.Q3V1 )
# plotting result
x.temp = seq(100,330,1)
glm.cf = coef(dat.Q3V1.glm)
y.temp = 1/(1+exp(-1*(glm.cf[1]+glm.cf[2]*x.temp)))
lines(x.temp,y.temp,col='red',lwd=2)

# Q4 example
plot(score~group,data=dat.Q4)
dat.Q4.G = lm(score~group,data=dat.Q4)

#
dat.Q4.NG1 = lm(score~income, data=dat.Q4)
dat.Q4.NG2 = lm(score~gender, data=dat.Q4)
dat.Q4.NG3 = lm(score~study, data=dat.Q4)
dat.Q4.NG4 = lm(score~height, data=dat.Q4)

dat.Q4.NG5 = lm(score~study+group,data=dat.Q4)
dat.Q4.NG6 = lm(score~study+income,data=dat.Q4)