In this section, we will do a linear regression of the building's energy efficiency measure, heating load (Y1) as a function of the building parameters. It would be useful to do a preliminary descriptive analysis to find which building variables are statistically significant. For this, we will first create bivariate plots of Y1 and all the X variables. We will also compute the Spearman correlation between Y1 and all the X variables. The R script for performing these tasks is as follows:

>library(ggplot2)
>library(gridExtra)

>df <- read.csv("ENB2012_data.csv",header = T)
>df <- df[,c(1:9)]
>str(df)
>df[,6] <- as.numeric(df[,6])
>df[,8] <- as.numeric(df[,8])

>attach(df)
>bp1 <- ggplot(data = df,aes(x = X1,y = Y1)) + geom_point()
>bp2 <- ggplot(data = df,aes(x = X2,y = Y1)) + geom_point()
>bp3 <- ggplot(data = df,aes(x = X3,y = Y1)) + geom_point()
>bp4 <- ggplot(data = df,aes...