1          OPTIONS NONOTES NOSTIMER NOSOURCE NOSYNTAXCHECK;
 61         
 62         /* mathlogreg4.sas */
 63         
 64         %include '/folders/myfolders/441s18/Lecture/mathread1.sas';
 NOTE: Format YNFMT is already on the library WORK.FORMATS.
 NOTE: Format YNFMT has been output.
 NOTE: Format CRSFMT is already on the library WORK.FORMATS.
 NOTE: Format CRSFMT has been output.
 NOTE: Format NFMT is already on the library WORK.FORMATS.
 NOTE: Format NFMT has been output.
 
 NOTE: PROCEDURE FORMAT used (Total process time):
       real time           0.00 seconds
       cpu time            0.00 seconds
       
 
 178        title2 'Use missing values for prediction';
 179        
 180        /* Make missing data indicators */
 
 NOTE: The infile '/folders/myfolders/441s18/Lecture/exploremath.data.txt' is:
       Filename=/folders/myfolders/441s18/Lecture/exploremath.data.txt,
       Owner Name=root,Group Name=vboxsf,
       Access Permission=-rwxrwx---,
       Last Modified=18Jan2016:17:34:49,
       File Size (bytes)=44583
 
 NOTE: 579 records were read from the infile '/folders/myfolders/441s18/Lecture/exploremath.data.txt'.
       The minimum record length was 75.
       The maximum record length was 75.
 NOTE: Missing values were generated as a result of performing an operation on missing values.
       Each place is given by: (Number of times) at (Line):(Column).
       99 at 101:24   99 at 102:18   
 NOTE: The data set WORK.MATHEX has 579 observations and 25 variables.
 NOTE: DATA statement used (Total process time):
       real time           0.01 seconds
       cpu time            0.01 seconds
       
 181        data mathex2;
 
 182             set mathex;
 183             if hsgpa+hscalc+precalc = . then missused = 1 ;
 184                else missused = 0;
 185             format missused ynfmt.;
 186             label missused   = 'Any of hsgpa hscalc precalc missing';
 187        
 
 NOTE: Missing values were generated as a result of performing an operation on missing values.
       Each place is given by: (Number of times) at (Line):(Column).
       142 at 183:14   204 at 183:21   
 NOTE: There were 579 observations read from the data set WORK.MATHEX.
 NOTE: The data set WORK.MATHEX2 has 579 observations and 26 variables.
 NOTE: DATA statement used (Total process time):
       real time           0.00 seconds
       cpu time            0.00 seconds
       
 188        proc freq data=mathex2;
 
 189             title2 'Percent passed for the record';
 190             tables passed;
 191        run;
 
 NOTE: There were 579 observations read from the data set WORK.MATHEX2.
 NOTE: PROCEDURE FREQ used (Total process time):
       real time           0.04 seconds
       cpu time            0.04 seconds
       
 
 192        
 193        proc freq data=mathex2;
 194             title2 'Missingness on variables used, as a predictor of passing';
 195             tables missused * passed / nocol nopercent chisq;
 196        run;
 
 NOTE: There were 579 observations read from the data set WORK.MATHEX2.
 NOTE: PROCEDURE FREQ used (Total process time):
       real time           0.05 seconds
       cpu time            0.05 seconds
       
 
 197        
 198        /* If missing on hsgpa, hscalc or precalc, give them an estimated
 199           probabilty of passing of 0.348. If not missing, use the model with
 200           hsgpa, hscalc and precalc to calculate the estimated probabilities. */
 201        
 202        proc logistic data = mathex2 noprint;
 203             title3 'Look at estimated probabilities';
 204             model passed (event='Yes') = hsgpa hscalc precalc;
 205             output out=mathex3 prob=pihat;
 206        /* The data set mathex3 has everything in mathex2, and also pihat */
 207        run;
 
 NOTE: PROC LOGISTIC is modeling the probability that passed='Yes'.
 NOTE: Convergence criterion (GCONV=1E-8) satisfied.
 NOTE: There were 579 observations read from the data set WORK.MATHEX2.
 NOTE: The data set WORK.MATHEX3 has 579 observations and 28 variables.
 NOTE: PROCEDURE LOGISTIC used (Total process time):
       real time           0.00 seconds
       cpu time            0.01 seconds
       
 
 208        
 209        proc print data=mathex3 (obs=13);
 210                             /* List only the first 13 observations */
 211             var hsgpa hscalc precalc pihat;
 212        run;
 
 NOTE: There were 13 observations read from the data set WORK.MATHEX3.
 NOTE: PROCEDURE PRINT used (Total process time):
       real time           0.03 seconds
       cpu time            0.03 seconds
       
 
 213        
 214        /* Based on invariance and the Law of Total Probability (double expectation),
 215           I predict that the mean pihat will be around 0.624, the proportion of
 216           students with non-missing data who passed. */
 217        
 218        proc univariate normal plot data=mathex3;
 219             title2 'Explore the distribution of estimated probabilities';
 220             where missused = 0; /* Should have n=375 */
 221             var pihat;
 222        run;
 
 NOTE: PROCEDURE UNIVARIATE used (Total process time):
       real time           0.66 seconds
       cpu time            0.22 seconds
       
 
 223        
 224        /* Where should we draw the line, and predict that the student will pass?
 225           In the exploratory sample, 52.68% passed overall, so find the
 226           pihat that cuts of the bottom 52.68%, and use that as a cutoff.*/
 227        
 228        data mathex4;
 229             set mathex3;
 230             if pihat = . then pihat = 0.348;
 231             /* Proportion with missing data who passed  */
 232        
 
 NOTE: There were 579 observations read from the data set WORK.MATHEX3.
 NOTE: The data set WORK.MATHEX4 has 579 observations and 28 variables.
 NOTE: DATA statement used (Total process time):
       real time           0.01 seconds
       cpu time            0.00 seconds
       
 233        proc freq data = mathex4;
 
 234             title2 'Seek the 52.68 percentile';
 235             tables pihat;
 236        run;
 
 NOTE: There were 579 observations read from the data set WORK.MATHEX4.
 NOTE: PROCEDURE FREQ used (Total process time):
       real time           0.45 seconds
       cpu time            0.41 seconds
       
 
 237        
 238        /* There was a pihat right at the 52.68th percentile: 0.4063743825.
 239           If pihat > 0.406, predict the student will pass. For the record and
 240           before peeking, I predict 75% correct. */
 241        
 242        /* Read and transform the replication data -- same code as mathread1.sas
 243           except no keep statement.  Creates the data set mathrep. */
 244        %include '/folders/myfolders/441s18/Lecture/readexplor.sas';
 NOTE: Format YNFMT is already on the library WORK.FORMATS.
 NOTE: Format YNFMT has been output.
 NOTE: Format CRSFMT is already on the library WORK.FORMATS.
 NOTE: Format CRSFMT has been output.
 NOTE: Format NFMT is already on the library WORK.FORMATS.
 NOTE: Format NFMT has been output.
 
 NOTE: PROCEDURE FORMAT used (Total process time):
       real time           0.00 seconds
       cpu time            0.00 seconds
       
 
 358        /* The data step continues ... */
 359        
 360        b0 = -14.7970; b1 = 0.1173; b2 = 0.0638; b3 = 0.2989; /* Exploratory sample values */
 361        L = b0 + b1*hsgpa + b2*hscalc + b3*precalc;
 362        pihat = exp(L)/(1+exp(L)); /* Some will be missing. */
 363        if pihat = . then pihat = 0.348; /* Sample proportion passed in the exploratory sample */
 364        if pihat > 0.406 then Prediction = 'Pass    '; else Prediction = 'Not Pass';
 365        
 366        /* Commented out
 367        proc freq data = mathrep;
 368             title2 'Checking Prediction';
 369             tables pihat * Prediction / norow nocol nopercent missing;
 370        run;
 371        */
 372        
 
 NOTE: The infile '/folders/myfolders/441s18/Lecture/replicmath.data.txt' is:
       Filename=/folders/myfolders/441s18/Lecture/replicmath.data.txt,
       Owner Name=root,Group Name=vboxsf,
       Access Permission=-rwxrwx---,
       Last Modified=04Feb2018:14:39:27,
       File Size (bytes)=38214
 
 NOTE: 579 records were read from the infile '/folders/myfolders/441s18/Lecture/replicmath.data.txt'.
       The minimum record length was 64.
       The maximum record length was 64.
 NOTE: Missing values were generated as a result of performing an operation on missing values.
       Each place is given by: (Number of times) at (Line):(Column).
       81 at 281:24    81 at 282:18    97 at 361:12    30 at 361:23    59 at 361:35    186 at 362:9    186 at 362:18
       186 at 362:19   
 NOTE: The data set WORK.MATHREP has 579 observations and 41 variables.
 NOTE: DATA statement used (Total process time):
       real time           0.00 seconds
       cpu time            0.01 seconds
       
 373        proc freq data = mathrep;
 
 374             title2 'How good is the prediction?';
 375             tables Prediction*passed / nocol;
 376        
 
 NOTE: There were 579 observations read from the data set WORK.MATHREP.
 NOTE: PROCEDURE FREQ used (Total process time):
       real time           0.02 seconds
       cpu time            0.02 seconds
       
 377        proc iml;
 
 NOTE: IML Ready
 378             PercentCorrect = 35.75 + 34.54;
 379             print PercentCorrect;
 380        run;
 NOTE: Module MAIN is undefined in IML; cannot be RUN.
 381        
 382        /* Prediction may be better at the extremes. */
 383        
 NOTE: Exiting IML.
 NOTE: PROCEDURE IML used (Total process time):
       real time           0.01 seconds
       cpu time            0.01 seconds
       
 384        proc freq data = mathrep;
 
 385             tables pihat*passed / norow nocol nopercent;
 386        run;
 
 NOTE: There were 579 observations read from the data set WORK.MATHREP.
 NOTE: PROCEDURE FREQ used (Total process time):
       real time           0.79 seconds
       cpu time            0.78 seconds
       
 
 387        
 388        
 389        
 390        
 391        
 392        
 393        OPTIONS NONOTES NOSTIMER NOSOURCE NOSYNTAXCHECK;
 406