2a) Download data on LOS AN GELES COUNTY ONLY from http://www.dqnews .com/Charts/MonthlyCharts/LATimesCharts/ZIPLAT.aspx . Or use the data from 2012 which are on the course website http://wwwstat .ucla.edu/~frederic/202a/F 13 in the le LAhousingprices.txt. Let Y = sales of single family homes in August, X1 = median price of a single family residence (SFR) in thousands of dollars, X2 = median price of a condo in thousands of dollars, and X3 : median home price per square foot, in dollars. (Be careful when loading the data from the website, because some of the names of the cities have spaces in them, and some of the numbers have commas in them. These have been corrected in the .txt le on my website.) Each of these 4 vectors initially has length 269. If any row has an "n/a" in it for any of these 4 variables, then remove this entire row. Now each vector will have length 217. 2b) Perform regression (with intercept) of Y on X: {X1, X2, X3} to compute a vector of parameter estimates, [3: (g, 131, g, [33), where [30 1s the estimated intercept and for 1: 1,2,3, is the slope corresponding to explanatory variable X. Record [31. 2c) Let i = 1. Perform regression with intercept of Y on X with row i removed from the dataset. Let )3\") denote your resulting vector of parameter estimates, so that 51"\" is your estimate of the slope corresponding to X1. Record [31\") - [31. 2d) Repeat step 2c) fori : 2, 3, 4, ..., 217. 2e) Plot the inuences, Z31\") - 231, versus i. That is, the xaxis will span from i = 1 to 217, and the y-axis will be 1'31\") - B1: which indicates the inuence of observation i on the estimated slope