# # script for topic 8a -- part 3 source("../gnrnd4.R") gnrnd4(365340906, 6340400805, 2200003) L1 L2 # main_hold <- "Problem for Topic 8a, part3" plot(L1,L2,main=main_hold, xlim=c(0,26), ylim=c(0,55), xaxp=c(0,26,13), yaxp=c(0,55,11), las=1, pch=16, col="red", mar=c(2,0,0,0)+0.2, ylab="y values", xlab="x values") abline(h=seq(0,55,5), v=seq(0,26,1), lty="dotted", col="blue") # get the two values for our regression equation, # the intercept and the slope lm( L2 ~ L1) # From the output of that command we see that the # intercept is 13.81 and the slope is 1.27. So # our regression equation is y = 13.81 + 1.27*x # We sill graph it on our plot abline( 13.81, 1.27, col="darkred") # # We can find the expected value when x=18 ev <- 13.81 + 1.27*18 ev # And then we can plot that point on the graph # with a green triangle points( 18, ev, pch=17, col="darkgreen") # And we can find the expected value for x=7 ev <- 13.81 + 1.27*7 ev # And then we can plot that point on the graph # with a blue square points( 7, ev, pch=15, col="blue") # # Now find the correlation coefficient cor( L1, L2 ) # # One other issue is to find some residual values. # Here we do that the hard way, we will compute the # observed - expected value for a given x value. # Find the residual value when x=17. The observed # value when x=17 is y=26. We need to compute the # expected value and then subtract that from 26 to # get the residual value. 26 - ( 13.81 + 1.27*17) # # If we find all of the # residual values and # then get a scatter plot # of the x and residual # values, we want to see # the points all over the # scatter plot. res_vals <- L2 - ( 13.81 + 1.27*L1) plot( L1, res_vals, main="Residuals")