Delete observations randomly based on percentage and impute them

[chintanpatel]

data RandClass(drop=i);
call streaminit(1234);
set sashelp.cars(keep=Cylinders Horsepower Enginesize);
array x {*} _numeric_;
do i = 1 to dim(x);
   if rand("Bern", 0.1) then         /* p=0.4 ==> about 40% missing */
      x[i]=.; end;
run;
proc print;
run;

data sashelp;
set sashelp.cars;
update cars Randclass;
by Cylinders Horsepower Enginesize;
run;
proc print cars;

I have removed data values randomly from three variables cylinders horsepower and enginesize from sashelp.cars. when i am trying to update newly created varaibles with missing value in data set sashelp.cars its giving me errors like

67 data sashelp;

68 set sashelp.cars;

69 update cars Randclass;

ERROR: File WORK.CARS.DATA does not exist.

70 by Cylinders Horsepower Enginesize;

71 run;

proc print cars;

____

22

202

ERROR 22-322: Syntax error, expecting one of the following: ;, BLANKLINE, CONTENTS, DATA, DOUBLE, GRANDTOTAL_LABEL, GRANDTOT_LABEL,

GRAND_LABEL, GTOTAL_LABEL, GTOT_LABEL, HEADING, LABEL, N, NOOBS, NOSUMLABEL, OBS, ROUND, ROWS, SPLIT, STYLE,

SUMLABEL, UNIFORM, WIDTH.

ERROR 202-322: The option or parameter is not recognized and will be ignored.

can anyone help me in this errors which is coming when i am trying to update. half part is working correct even i am getting as an output randclass dataset with 3 variables. in the 2nd part when i am trying to update its giving me errors.

[Reeza]

The error is clear, do you have a dataset CARS created, check your library? If not then check your code. I think you meant to create a dataset called CARS but you called it SASHELP instead? Make sure you're consistent with the names used. 

[art297]

Your code won't work for a number of reasons. For one you select every cell, as the random number will always be greater than 0, thus the condition will always be true.

Second, you never create the file cars but, rather, a file called sashelp

Art, CEO, AnalystFinder.com