quietly {
/* Calculate Bivariate Regression pp. 50-53, Section 5-5-3 */
 log using regsol, smcl replace
 noisily di _n "Regression Solution Using the Long-hand method (pp. 48-50)"
 sum age                                             /*************************************/
 gen xdev=age-r(mean)                                /* Deviation of X from the Mean of X */
 sum sbp                                             /*                                   */
 gen ydev=sbp-r(mean)                                /* Deviation of X from the Mean of X */
 gen xprod=xdev*ydev                                 /* Crossproduct of X and Y           */
 gen x2=xdev^2                                       /* Deviation of X-squared            */
 sum xprod                                           /* Calculate summary statistics      */
 local num=r(sum)                                    /*                                   */
 sum x2                                              /*                                   */
 local denom=r(sum)                                  /*                                   */
 local beta_1=round(`num'/`denom',.01)               /* Calculate slope coefficient       */
 sum sbp                                             /*                                   */
 local meany=r(mean)                                 /*                                   */
 sum age                                             /*                                   */
 local meanx=r(mean)                                 /*                                   */
 local beta_0=round(`meany'-(`meanx'*`beta_1'),.02)  /* Calculate intercept coefficient   */
 noisily di _n "Y-hat= "`beta_0' "+" `beta_1'        /* Display results                   */
                                                     /*************************************/


 tempvar xy x2                                       /*************************************/
 gen xy=age*sbp                                      /* Calculate Regression using other  */
 gen age2=age^2                                      /*   formula                         */
 format sbp age xy age2 %9.0gc                       /*************************************/
 format xdev ydev xprod x2 %9.2fc
 noisily list, noobs sum(age sbp xy x2)              
 noisily ret list

 noisily regress sbp age                             /* Regression using Stata procedure  */
 drop xdev- age2
 log close
 view regsol.smcl
}