Skip to content

Commit

Permalink
latest sas code
Browse files Browse the repository at this point in the history
  • Loading branch information
@pradeep17j
pradeep17j committed Aug 11, 2019
1 parent 7b5f4d5 commit 3a091c8
Showing 1 changed file with 2 additions and 265 deletions.
267 changes: 2 additions & 265 deletions AppliedStats_Proj2/Project2.sas
View file
Original file line number Diff line number Diff line change
Expand Up @@ -20,294 +20,31 @@ PROC CONTENTS DATA=kobeshot; RUN;
%web_open_table(kobeshot);


%web_drop_table(testkobeshot);

%web_drop_table(new_data);


FILENAME REFFILE '/home/pradeep17j0/PradeepFolder/MSDS6372/Proj2/test_KobeData.csv';

PROC IMPORT DATAFILE=REFFILE
DBMS=CSV
OUT=testkobeshot;
GETNAMES=YES;
RUN;

PROC CONTENTS DATA=testkobeshot; RUN;


%web_open_table(testkobeshot);


proc print data=kobeshot(obs=10);
run;

proc print data=testkobeshot(obs=10);
run;


proc freq data=kobeshot;
table shot_made_flag shot_distance shot_made_flag*shot_distance / out=p outpct;
run;


data kobeshot;
set kobeshot;
if (shot_zone_range = 'Less Than 8 ft.') then szone=0;
if (shot_zone_range ='8-16 ft.') then szone=1;
if (shot_zone_range = '16-24 ft.') then szone=2;
if (shot_zone_range = '24+ ft.') then szone=3;
if (shot_zone_range = 'Back Court Shot') then szone=4;

;

proc freq data=kobeshot;
table shot_made_flag szone shot_made_flag*szone / out=p1 outpct;
run;

proc print data=p(obs=10);
run;

proc sgplot data=p;
scatter x=shot_distance y=pct_col/ group=shot_made_flag;
run;

proc sgplot data=p1;
by shot_made_flag;
scatter x=szone y=pct_col;
run;



data logit;
set p;
logit = log(pct_col/100/(1-pct_col/100));
;

proc sgplot data=logit;
by shot_made_flag;
scatter x=shot_distance y=logit;
scatter x=shot_distance y=logit/ group=shot_made_flag;
run;

/* "16-24 ft." "24+ ft." "8-16 ft." "Back Court Shot" "Less Than 8 ft." */

data kobeshot;
set kobeshot;
if (shot_zone_range == 'Less Than 8 ft.') then szone=0;
if (shot_zone_range == '8-16 ft.') then szone=1;
if (shot_zone_range == '16-24 ft.') then szone=2;
if (shot_zone_range == '24+ ft.') then szone=3;
if (shot_zone_range == 'Back Court Shot') then szone=4;

;



proc logistic data=kobeshot;
model shot_made_flag=shot_distance arena_temp loc_x loc_y ;
model shot_made_flag=shot_distance;
run;

data new_data;
set testkobeshot;
*shot_made_flag =.;
;


data combined_data;
set kobeshot new_data;
log_shotdist=log(shot_distance)
;

data tmp_kobedata;
set kobeshot;
n=ranuni(8);
proc sort data=tmp_kobedata;
by n;
data training testing;
set tmp_kobedata nobs=nobs;
if _n_<=.7*nobs then output training;
else output testing;
run;
*%web_drop_table(training);
*%web_drop_table(testing);
%web_drop_table(tmp_kobedata);




ods graphics on;
proc logistic data=kobeshot plots(only)=(oddsratio(range=clip)) descending outmodel=Model1 PLOTS(MAXPOINTS=NONE);
class shot_made_flag playoffs combined_shot_type shot_zone_range action_type;
model shot_made_flag(event='1') = shot_zone_range combined_shot_type playoffs arena_temp action_type
/ selection=STEPWISE cl expb lackfit ctable outroc=troc;
oddsratio shot_made_flag;
output out=predicted predicted=l;
score data=kobeshot outroc=vroc;

run;
ods graphics off;

ods graphics on;
proc logistic data=training descending outmodel=Model1 PLOTS(MAXPOINTS=NONE);
class shot_made_flag playoffs combined_shot_type shot_zone_range action_type;
model shot_made_flag(event='1') = shot_zone_range combined_shot_type playoffs arena_temp action_type loc_x
/ selection=STEPWISE cl expb lackfit ctable outroc=troc;
oddsratio shot_made_flag;
output out=predicted predicted=l;
score data=testing outroc=vroc fitstat ;
roc;

run;


proc logistic data=training;
class shot_made_flag playoffs combined_shot_type shot_zone_range action_type;
model shot_made_flag(event='1') = playoffs combined_shot_type shot_zone_range action_type
seconds_remaining avgnoisedb seconds_remaining*avgnoisedb
/ selection=STEPWISE cl expb lackfit ctable;
score data=testing fitstat ;
run;

/* final model */
proc logistic data=training;
class shot_made_flag playoffs(ref='0') combined_shot_type shot_zone_range(ref='Less Than 8 ft.')
action_type opponent matchup / param=reference;
model shot_made_flag(event='1') = playoffs action_type shot_zone_range
seconds_remaining avgnoisedb playoffs*shot_zone_range
/ selection=STEPWISE cl expb lackfit ctable outroc=troc stopres pprob=0.45 ;
score data=testing fitstat outroc=vroc ;
score data=testkobeshot out=results;
run;


/* for distance vs shot made */
proc logistic data=training;
class shot_made_flag playoffs combined_shot_type shot_zone_range(ref='Less Than 8 ft.')
action_type opponent matchup / param=reference;
model shot_made_flag(event='1') = shot_zone_range playoffs*shot_zone_range
/ selection=STEPWISE cl expb lackfit ctable ;
score data=testing fitstat ;
score data=testkobeshot out=results;
run;


proc discrim data=kobeshot pool=yes crossvalidate
testdata=testkobeshot testout=dataout crosslisterr;
class shot_made_flag;
var ;
run;



proc print data=dataout;
run;


ods graphics on;
proc logistic data=training descending outmodel=Model1 PLOTS(MAXPOINTS=NONE);
class shot_made_flag playoffs combined_shot_type shot_zone_range shot_zone_basic action_type shot_zone_area;
model shot_made_flag(event='1') = shot_zone_range combined_shot_type playoffs arena_temp action_type shot_zone_basic
shot_zone_area
/ selection=STEPWISE cl expb lackfit ctable outroc=troc;

output out=predicted predicted=l;
score data=testing outroc=vroc;

run;

/*
ods graphics on;
proc logistic data=combined_data plots(only)=(oddsratio(range=clip)) descending outmodel=Model1;
class shot_made_flag (ref='0') playoffs (ref='0') combined_shot_type /param=reference;
model shot_made_flag= shot_distance | arena_temp | loc_x |loc_y | playoffs |seconds_remaining
/ selection=STEPWISE cl expb lackfit ctable;
oddsratio shot_made_flag;
output out=predicted predicted=l;
run;
ods graphics off;
*/

ods graphics on;
proc logistic inmodel=Model1;
score data=combined_data fitstat clm outroc=vroc;
roc;
run;
ods graphics off;

title 'Discriminant Analysis of Remote Sensing Data on Five Crops';

data crops;
input Crop $ 1-10 x1-x4 xvalues $ 11-21;
datalines;
Corn 16 27 31 33
Corn 15 23 30 30
Soybeans 24 24 25 32
Soybeans 21 25 23 24
Corn 12 15 16 73
Soybeans 20 23 23 25
Soybeans 24 24 25 32
Soybeans 21 25 23 24
Soybeans 27 45 24 12
Soybeans 12 13 15 42
Soybeans 22 32 31 43
Cotton 31 32 33 34
Cotton 29 24 26 28
Cotton 34 32 28 45
Cotton 26 25 23 24
Cotton 53 48 75 26
Cotton 34 35 25 78
Sugarbeets22 23 25 42
Sugarbeets25 25 24 26
Sugarbeets34 25 16 52
Sugarbeets54 23 21 54
Sugarbeets25 43 32 15
Sugarbeets26 54 2 54
Clover 12 45 32 54
Clover 24 58 25 34
Clover 87 54 61 21
Clover 51 31 31 16
Clover 96 48 54 62
Clover 31 31 11 11
Clover 56 13 13 71
Clover 32 13 27 32
Clover 36 26 54 32
Clover 53 08 06 54
Clover 32 32 62 16

;

title2 'Using the Linear Discriminant Function';

proc discrim data=crops outstat=cropstat method=normal pool=yes
list crossvalidate crosslisterr listerr;
class Crop;
priors prop;
id xvalues;
var x1-x4;
run;


data test;
input Crop $ 1-10 x1-x4 xvalues $ 11-21;
datalines;
Corn 16 27 31 33
Soybeans 21 25 23 24
Cotton 29 24 26 28
Sugarbeets54 23 21 54
Clover 32 32 62 16
;title2 'Classification of Test Data';

proc discrim data=cropstat testdata=test testout=tout testlist listerr ;
class Crop;
testid xvalues;
var x1-x4;
run;

proc print data=tout;
title 'Discriminant Analysis of Remote Sensing Data on Five Crops';
title2 'Output Classification Results of Test Data';
run;



0 comments on commit 3a091c8

Please sign in to comment.