/* Final Part 1 SAS - STATS 381 */
options ls = 80 nocenter nodate ps=55 nonumber;
title1 'MATH 381 Spring 2018';
title2 'Final #1';
title3 'Billy ODonnell';

/* Establish data variable */
data defective;
/* Open an input file */
infile 'C:\Users\odonnelw\Desktop\QUALITY1_S18.dat';
/* Create an input variable */
input defectivechips;
/* Change title appropriate for this Q */
title4 'Question 1';
/* Print the output */
proc print noobs;
title4 'Question 1 - i';
/* Run our FREQ procedure, using our input variable */
proc freq;
    tables defectivechips;
title4 'Question 1 - v';
/* Run our means procedure to get the average number of chips per shipment */
proc means sum mean;

/* New data variable for next question */
data icecream;
infile 'C:\Users\odonnelw\Desktop\icecream1_s18.dat' firstobs = 2;

/* Run a nested do loop to read in our data
and establish brand classifications using if statements */
do brand = 1 to 3;
if brand = 1 then preference =      'Breyers ';
else if brand = 2 then preference = 'Dreyers ';
else preference = 'PrivateSelection';

/* Establish age classifications */
do age = 1 to 3;
    if age = 1 then AgeBracket = '< 21';
    else if age = 2 then AgeBracket = '21 - 45';
    else AgeBracket = ' > 45';
/* Input variable */
input ages @@;
output;
end;
end;
title4 'Question 2 - i';
proc print noobs;

title4 'Question 2 - ii';
/* FREQ Procedure for chi sq test for independence */
proc freq order = data;
    weight ages;
    tables age*brand/chisq expected nopercent norow nocol;

/* New data variable for next question */
data basketball;
infile 'C:\Users\odonnelw\Desktop\basketballs1_s18.dat' firstobs = 2;

/* Nested do loops for Defective v Non Defective */
do itemClassification = 1 to 2;
if itemClassification = 1 then entry = 'NonDefective';
else entry = 'Defective';

/* Secondary loop to classify by shift type */
do shift = 1 to 2;
    if shift = 1 then shiftname = 'Day';
    else shiftname = 'Nite';
input basketballs @@;
output;
end;
end;
title4 'Question 3 - i and ii';
proc print noobs;
title4 'Question 3 - iii';
/* Chi Square test for independence using FREQ procedure */
proc freq order = data;
    weight basketballs;
    tables itemClassification*shift/chisq expected nopercent norow nocol;

/* New data variable for next Q */
data ink;
/* Our input file has both colors, a data value and a proportion,
designate this (text value) by using $ */
input ink_color $ observed prop;
n = 80;
/* Create our expected number by multiplying N by proportion expected */
expected = n*prop;
/* Get our ChiSquare value */
chisq1 = (observed - expected)**2/expected;
datalines;
Black 28 .30
Blue 26 .30
Red 18 .25
Other 8 .15
;
title4 'Question 4';
proc print noobs;
/* Test Hypothesis */
proc means sum noprint;
    var chisq1;
    output out = a1 sum = chisq_sum;
proc print noobs;
/* New data variable to hold pvalue */
data pvalue1;
set a1;
pvalue = 1 - probchi(chisq_sum, 3);
proc print noobs;

/* New data value to do GoodnessOfFit */
data ink2;
input ink_color $ observed prop;
n = 80;
expected = n*prop;
chisq1 = (observed - expected)**2/expected;
datalines;
Black 28 .30
Blue 26 .30
Red 18 .25
Other 8 .15
;
proc print noobs;
title4 'Question 4';
/* Use Univariate Procedure to get our GoodnessOfFit
information, use estimated values because we have no
mean and sigma */
proc univariate;
    ods select GoodnessOfFit;
    var chisq1;
    histogram / normal(mu=est sigma = est);
proc print noobs;

title4 'Question 5';
/* New data var for next Q */
data lightbulbs;
infile 'C:\Users\odonnelw\Desktop\LIFETEST1_S18.dat' firstobs = 2;
input life @@;
proc print noobs;
/* Univariate to test for exponential distribution */
proc univariate;
    ods select GoodnessOfFit;
    var life;
    histogram/exponential(sigma=est);


/* New data var for next Q */
data exams;
infile 'C:\Users\odonnelw\Desktop\COMPLETION1_S18.dat' firstobs = 2;
input completion @@;
title4 'Question 6';
proc print noobs;
/* Univariate to test for normal distribution using mean = 95 minutes
and sigma = 15 minutes */
proc univariate;
    ods select GoodnessOfFit;
    var completion;
    histogram/normal(mu=95 sigma=15);
run; quit;