INDEX creation using DATA step

[yadaw]

Hi All,

I have a DATA SET called FR that has variables like: stu_id, name, age, gender. I want to create a simple index on stu_id using DATA step (Yes, we can do the same by PROC SQL or PROC DATASETS but here I want to create using DATA step).

I have written the below program but unable to create index:

DATA FR1 (INDEX=(stu_id=(stu_id)));

       SET FR;

RUN;

Can anybody help me on this?

And also, Is there any way to DELETE the created index by DATA step (I don't want to DELETE using PROC SQL or PROC DATASETS) ?

Thank You,

Yadaw

[yadaw]

Hi All,

I have a DATA SET called FR that has variables like: stu_id, name, age, gender. I want to create a simple index on stu_id using DATA step (Yes, we can do the same by PROC SQL or PROC DATASETS but here I want to create using DATA step).

I have written the below program but unable to create index:

DATA FR1 (INDEX=(stu_id=(stu_id)));

       SET FR;

RUN;

Can anybody help me on this?

And also, Is there any way to DELETE the created index by DATA step (I don't want to DELETE using PROC SQL or PROC DATASETS) ?

Thank You,

Yadaw

[yadaw]

Hi All,

I have a DATA SET called FR that has variables like: stu_id, name, age, gender. I want to create a simple index on stu_id using DATA step (Yes, we can do the same by PROC SQL or PROC DATASETS but here I want to create using DATA step).

I have written the below program but unable to create index:

DATA FR1 (INDEX=(stu_id=(stu_id)));

       SET FR;

RUN;

Can anybody help me on this?

And also, Is there any way to DELETE the created index by DATA step (I don't want to DELETE using PROC SQL or PROC DATASETS) ?

Thank You,

Yadaw

[Kurt_Bremser]

The tool for creating an index on an existing datset is PROC DATASETS.

[Ksharp · Accepted solution]

I could build an index, but don't know how to delete it yet.

data class(index=(name));
 set sashelp.class;
run;
proc contents data=class;
run;

Composite index.

data class(index=(comp=(name age)));
 set sashelp.class;
run;
proc contents data=class;
run;

[Ksharp]

You could just simple copy the table to eliminate the index.

data x;
set class;
run;
proc contents data=x;
run;

[RichardDeVen]

A data set index can be created in different ways

• As specified by an output data set with option INDEX=
  • Simple indices (single columns) can not be named something else
  • Compound indices are specified <index-name>=(<column1> ... <column-n>)
• PROC DATASETS; MODIFY
• PROC SQL; CREATE INDEX

Index management is done with either DATASETS or SQL.  Read the documentation for more information about uniqueness and non-null indices and other data index topics such as integrity constraints and foreign keys.

Examples:

data cars(index=(Make Model type_drive=(Type DriveTrain)));
  set sashelp.cars;
run;

proc datasets nolist lib=work; 
  modify cars;
  index delete Make Model;
  index create Origin;
run;

proc sql;
  create index Make on work.cars;
  drop index type_drive from work.cars;
  create index type_cyl on work.cars(Type, Cylinders) ;

NOTE: SAS data sets do not have RDBMS features such as triggers or sequence numbers.