Finishing up with the Bang

. infile veloc dist using case0701.asc
'velocity' cannot be read as a number for veloc[1]
'distance' cannot be read as a number for dist[1]
(25 observations read)

. regress dist veloc, nocons
  Source |       SS       df       MS                  Number of obs =      24
---------+------------------------------               F(  1,    23) =  101.14
   Model |   24.048893     1   24.048893               Prob > F      =  0.0000
Residual |  5.46890204    23  .237778349               R-squared     =  0.8147
---------+------------------------------               Adj R-squared =  0.8067
   Total |   29.517795    24  1.22990813               Root MSE      =  .48763
------------------------------------------------------------------------------
    dist |      Coef.   Std. Err.       t     P>|t|       [95% Conf. Interval]
---------+--------------------------------------------------------------------
   veloc |   .0019218   .0001911     10.057   0.000       .0015265    .0023171
------------------------------------------------------------------------------
As derived in class, the conversion from megaparsecs/km/sec to years (in billions) is obtained by multiplying the regression coefficient by the product of 3.26 (the number of light-years per parsec) and 299.792 (the number of kilometers in one light-second x 10^{-3}). 
. display 3.26 * 299.792 *  .0019218  
1.8782173

Designing a Screening Test (AFP)

. clear

. use afpnorm

. count
 1411

. summarize afp
Variable |     Obs        Mean   Std. Dev.       Min        Max
---------+-----------------------------------------------------
     afp |    1407    40.46923   19.85923   4.669257    349.324  

. count if afp>80
   50

. di 50/1411
 .03543586

graph afp, histogram saving(a) 

. graph afp, histogram saving(a)

graph afp, histogram bin(50) saving(b) 

. graph afp, histogram bin(50) saving(b)

graph afp, oneway saving(c) 

. graph afp, oneway saving(c)

graph afp, oneway jitter(2) saving(d) 

. graph afp, oneway jitter(2) saving(d)

graph afp, oneway jitter(2) box saving(f) 

. graph afp, oneway jitter(2) box saving(f)

. summarize afp, detail
                      Alpha-fetoprotein
-------------------------------------------------------------
      Percentiles      Smallest
 1%     13.17077       4.669257
 5%     19.08685       6.820958
10%     22.53343       7.078061       Obs                1407
25%     28.61698       8.908448       Sum of Wgt.        1407
50%     36.78168                      Mean           40.46923
                        Largest       Std. Dev.      19.85923
75%      48.6183       136.5922
90%      61.2522        153.546       Variance        394.389
95%     70.59787       242.2572       Skewness       4.434543
99%     106.2718        349.324       Kurtosis       53.67288

. lookup percentile
[R]     centile  . . . . . . . . . . .  Report centile and confidence interval
        (help centile)
[R]     codebook . . . . .  Produce a codebook describing the contents of data
        (help codebook)
[R]     collapse . . . . . . . . . . . .  Make dataset of means, medians, etc.
        (help collapse)
[R]     diagplots  . . . . . . . . . . . . . . Distributional diagnostic plots
        (help diagplots)
[R]     egen . . . . . . . . . . . . . . . . . . . . .  Extensions to generate
        (help egen)
[R]     lv . . . . . . . . . . . . . . . . . . . . . . . Letter-value displays
        (help lv)
[R]     pctile . . . . . . . . . . . .  Create variable containing percentiles
        (help pctile)
[R]     qreg . . . . . . . . . . . . .  Quantile (including median) regression
        (help qreg)
[R]     summarize  . . . . . . . . . . . . . . . . . . . .  Summary statistics
        (help summarize)
[R]     table  . . . . . . . . . . . . . . . . .  Tables of summary statistics
        (help table)

. help centile
[ OUTPUT FROM THE help COMMAND has been omitted here ]

. centile afp, centile(50 75 95 97.5)
                                                   -- Binom. Interp. --
Variable |     Obs  Percentile      Centile        [95% Conf. Interval]
---------+-------------------------------------------------------------
     afp |    1407         50      36.78168        35.61309    37.56226
         |                 75       48.6183        47.07178    49.95195
         |                 95       70.8102        67.62652    75.41453
         |               97.5      83.78058        78.91603    89.77014

graph afp gestwk, saving(g) 

. graph afp gestwk, saving(g)

. drop if gestwk<=14
(3 observations deleted)

. drop if gestwk>20
(3 observations deleted)
In order for the graph command to produce different figures for each value of a grouping variable, the data set needs to be sorted in order of that grouping variable first. 
. sort gestwk

graph afp, box by(gestwk) saving(h) 

. graph afp, box by(gestwk) saving(h)

graph lafp, box by(gestwk) saving(j) 

. graph lafp, box by(gestwk) saving(j)

. regr lafp gestwk
  Source |       SS       df       MS                  Number of obs =    1401
---------+------------------------------               F(  1,  1399) =  211.20
   Model |  31.8475509     1  31.8475509               Prob > F      =  0.0000
Residual |  210.963012  1399  .150795577               R-squared     =  0.1312
---------+------------------------------               Adj R-squared =  0.1305
   Total |  242.810563  1400  .173436116               Root MSE      =  .38832
------------------------------------------------------------------------------
    lafp |      Coef.   Std. Err.       t     P>|t|       [95% Conf. Interval]
---------+--------------------------------------------------------------------
  gestwk |   .1235966   .0085048     14.533   0.000       .1069131    .1402801
   _cons |   1.552031   .1420225     10.928   0.000       1.273431    1.830631
------------------------------------------------------------------------------

. predict yhat

. gen cutoff = yhat + 1.96* 0.38832

. sort gestday

graph lafp cutoff gestday, s(oi) c(.l) saving(k) 

. graph lafp cutoff gestday, s(oi) c(.l) saving(k)