ALT与骨密度SAS代码

/*基本信息表身高，体重，年龄，性别 */

options validvarname=any;/*导入数据，确定可以使用中文变量 */
proc import datafile = "C:\Users\GaryHo\Desktop
\LJ_jibenxinxi.xlsx" dbms = excel out = test;
sheet = "sheet1";
run;

data test;
set test;
drop 总检日期  体检编号  出生日期    总检日期    是否有问卷数据    是否有体检数据    是否有骨密度数据;性别 /*删除不要的变量列 只保留ID  年龄   身高    体重 */   
run;

proc export data =test /*导出数据*/
outfile = "C:\Users\GaryHo\Desktop\SAS_jibenxinxi.xlsx"
dbms = xlsx replace;
sheet = 'sheet1' ;
run;

/*髋骨密度表 */

options validvarname=any;
proc import datafile ="C:\Users\GaryHo\Desktop\HipBMD.xlsx"
dbms = excel out = test;sheet = "sheet1";
run;

proc sql;
create table test1 as select 体检编号,TOT_BMD,Hip_T from test where 体检编号 >600000000000;
quit;

proc export data =test1 /*导出数据*/
outfile = "C:\Users\GaryHo\Desktop\SASHipBMD.xlsx"
dbms = xlsx replace;
sheet = 'sheet1' ;
run;


/*腰椎骨密度表 */

options validvarname=any;
proc import datafile ="C:\Users\GaryHo\Desktop\LSBMD.xlsx"
dbms = excel out = test;sheet = "sheet1";
run;

proc sql;
create table test1 as select 体检编号,TOT_BMD,SPINE_T from test where 体检编号 >600000000000;
quit;

proc export data =test1 /*导出数据*/
outfile = "C:\Users\GaryHo\Desktop\SAS_LSBMD.xlsx"
dbms = xlsx replace;
sheet = 'sheet1' ;
run;



/*五张表内连结，所有需要用的信息已经有了。现在开始数据处理清洗部分*/
proc sql;
creat table test1 as
select * from SAS_LSBMD1 as a inner join  
Sas_shenghua1 as b on a.ID=b.ID
inner join  
SAS_1 as c on a.ID=c.ID
inner join  
Sas_jibenxinxi1 as d on d.ID=a.ID;
inner join jibingwenjuan1 as e on e.id=a.id;
quit;

proc export data =test1 /*导出数据*/
outfile = "C:\Users\GaryHo\Desktop\zhenglihao.xlsx"
dbms = xlsx replace;
sheet = 'sheet1' ;
run;

数值型数据清理：

/*删除重复值，可以借鉴笔记 SAS重复值部分/



/*用freq统计某变量的值重复的情况，并且可以用来检查*/
proc freq data = test noprint ;
    table ID/  out =Nodups6 (keep = ID Count where = (Count > 1)) ;
run ;

proc sort data = test  out=_unique dupout = _dups 
nodupkey ;
by ID;
run ;

/*检查,如何没有显示即成功*/
proc freq data = _unique noprint ;
    table ID/  out =Nodups6 (keep = ID Count where = (Count > 1)) ;
run ;



/*missing缺失值处理*/

data missing;
set test;
array a _numeric_;
do over a;
if missing(a) then output;
end;
array b _character_;
do over b;
if missing(b) then output;
end;
run;



/*与ALT，AST偏相关情况 */
PROC corr pearson  ;
var ALT AST age sex height weight    BMI wc hipline    WHR HTOT_BMD LS_BMD;
run;


/*与ALT，AST偏相关情况，校正男女 */
PROC corr pearson  ;
var ALT AST  height weight    BMI wc hipline    WHR HTOT_BMD LS_BMD;
PARTIAL age sex;
run;

/*与ALT，AST偏相关情况 */
PROC corr pearson  ;
var ALT AST age sex height weight    BMI wc hipline    WHR HTOT_BMD LS_BMD;
run;


/*与ALT，AST偏相关情况，校正男女 */
PROC corr pearson  ;
var ALT AST  height weight    BMI wc hipline    WHR HTOT_BMD LS_BMD;
PARTIAL age sex;
run;

/*分男女，与ALT，AST与hip BMD 回归模型 */

proc sort data=test;
by sex;
run;

proc reg;
model HTOT_BMD=ALT  age  /stb ;
run;
proc reg;
model HTOT_BMD=ALT WC  age  /stb ;
run;
proc reg;
model HTOT_BMD=ALT BMi  age  /stb ;
run;
proc reg;
model HTOT_BMD=ALT BMi WC age  /stb ;
run;

/*分男女，与ALT，AST与LS BMD 回归模型 */
proc reg;
model LS_BMD=ALT  age  /stb ;
run;
proc reg;
model LS_BMD=ALT WC  age  /stb ;
run;
proc reg;
model LS_BMD=ALT BMi  age  /stb ;
run;
proc reg;
model LS_BMD=ALT BMi WC  age  /stb vif collin;
run;

/*将BMI分组，
与ALT，AST与LS BMD 回归模型 */

proc sort data=test;
by BMIgroup;
run;


proc reg;
model HTOT_BMD=ALT sex age  /stb ;
by BMIgroup;
run;

proc reg;
model HTOT_BMD=ALT sex age wc /stb ;
by BMIgroup;
run;

proc reg;
model LS_BMD=ALT sex age  /stb ;
by BMIgroup;
run;

proc reg;
model LS_BMD=ALT sex age  wc /stb ;
by BMIgroup;
run;


/*将BMI分组，男女
与ALT，AST与LS BMD 回归模型 */

proc sql;
create table test_women as
select * from test where sex=0;
quit;


proc reg;
model HTOT_BMD=ALT  age  /stb ;
by BMIgroup;
run;

proc reg;
model HTOT_BMD=ALT  age wc /stb ;
by BMIgroup;
run;

proc reg;
model LS_BMD=ALT  age  /stb ;
by BMIgroup;
run;

proc reg;
model LS_BMD=ALT  age  wc /stb ;
by BMIgroup;
run;