topic Re: Error "Expecting a Name" in SAS Programming

Error "Expecting a Name"

JBerry — Sat, 10 Jan 2015 00:29:25 GMT

I am receiving a very frustrating SAS error when trying to use PROC PANEL."Expecting a name"

I have created some sample code that reproduces the error. Any ideas on what I'm missing?

876 PROC SORT DATA=TEMP OUT=TEMPSORTED;

877 BY OBS_DATE;

878 RUN;

NOTE: There were 26 observations read from the data set WORK.TEMP.

NOTE: The data set WORK.TEMPSORTED has 26 observations and 2 variables.

NOTE: PROCEDURE SORT used (Total process time):

real time 0.02 seconds

cpu time 0.01 seconds

879

880 PROC PANEL DATA=TEMPSORTED;

881 ID OBS_DATE;

ERROR 22-322: Expecting a name.

882 LAG DEGREES_F(1) / OUT=TEMP_W_LAG;

ERROR: Variable NAME not found.

883 RUN;

NOTE: The SAS System stopped processing this step because of errors.

NOTE: PROCEDURE PANEL used (Total process time):

real time 0.06 seconds

cpu time 0.00 seconds

Here is an entire section of code which reproduces the error:

DATA TEMP;

INPUT OBS_DATE mmddyy10. DEGREES_F @17;

FORMAT OBS_DATE mmddyy10.;

datalines;

11/01/2014 44

11/02/2014 53

11/03/2014 64

11/04/2014 61

11/05/2014 63

11/06/2014 52

11/07/2014 45

11/08/2014 49

11/09/2014 53

11/10/2014 65

11/11/2014 61

11/12/2014 33

11/13/2014 31

11/14/2014 29

11/15/2014 33

11/16/2014 33

11/17/2014 25

11/18/2014 21

11/19/2014 33

11/20/2014 30

11/21/2014 36

11/22/2014 54

11/23/2014 54

11/24/2014 51

11/25/2014 30

11/26/2014 32

;

PROC SORT DATA=TEMP OUT=TEMPSORTED;

BY OBS_DATE;

RUN;

;

PROC PANEL DATA=TEMPSORTED;

ID ;

LAG DEGREES_F(1) / OUT=TEMP_W_LAG;

RUN;

Re: Error "Expecting a Name"

JBerry — Sat, 10 Jan 2015 00:52:51 GMT

I actually figured it out - Proc Panel expects there to be at least 1 category to group by. This is solved by faking it like so:

(I'll post a new thread on ideas on how to do this more eloquently)

DATA TEMP;

INPUT OBS_DATE mmddyy10. DEGREES_F @17;

FORMAT OBS_DATE mmddyy10.;

datalines;

11/01/2014 44

11/02/2014 53

11/03/2014 64

11/04/2014 61

11/05/2014 63

11/06/2014 52

11/07/2014 45

11/08/2014 49

11/09/2014 53

11/10/2014 65

11/11/2014 61

11/12/2014 33

11/13/2014 31

11/14/2014 29

11/15/2014 33

11/16/2014 33

11/17/2014 25

11/18/2014 21

11/19/2014 33

11/20/2014 30

11/21/2014 36

11/22/2014 54

11/23/2014 54

11/24/2014 51

11/25/2014 30

11/26/2014 32

;

/* MAKE A FAKE COPY OF THE DATA AND APPEND TOGETHER INTO 1 DATASET */

DATA TEMP; SET TEMP;

FAKEGROUP=0;

RUN;

DATA TEMP_FAKE; SET TEMP;

FAKEGROUP=1;

RUN;

PROC APPEND BASE=TEMP DATA=TEMP_FAKE;

RUN;

/* SORT THE DATA BY THE FAKEGROUP AND OBS DATE */

PROC SORT DATA=TEMP OUT=TEMPSORTED;

BY FAKEGROUP OBS_DATE;

RUN;

/* USE PROC PANEL TO GENERATE THE LAG VARIABLES */

PROC PANEL DATA=TEMPSORTED;

ID FAKEGROUP OBS_DATE;

LAG DEGREES_F(1 2 3 4 5 6 7 8 9 10 11 12) / OUT=TEMP_W_LAG;

RUN;

/* DELETE THE FAKEGROUP SO YOU ARE BACK TO THE ORIGINAL DATASET */

DATA TEMP_W_LAG; SET TEMP_W_LAG(WHERE=(FAKEGROUP=0));

RUN;

Re: Error "Expecting a Name"

Patrick — Sat, 10 Jan 2015 01:39:42 GMT

For your special case you would also get away with a simple data step like below:

DATA TEMP;

INPUT OBS_DATE mmddyy10. DEGREES_F @17;

FORMAT OBS_DATE mmddyy10.;

datalines;

11/01/2014 44

11/02/2014 53

11/03/2014 64

11/04/2014 61

11/05/2014 63

11/06/2014 52

11/07/2014 45

11/08/2014 49

11/09/2014 53

11/10/2014 65

11/11/2014 61

11/12/2014 33

11/13/2014 31

11/14/2014 29

11/15/2014 33

11/16/2014 33

11/17/2014 25

11/18/2014 21

11/19/2014 33

11/20/2014 30

11/21/2014 36

11/22/2014 54

11/23/2014 54

11/24/2014 51

11/25/2014 30

11/26/2014 32

;

PROC SORT DATA=TEMP OUT=TEMPSORTED;

BY OBS_DATE;

RUN;

data TEMP_W_LAG(drop=_:);

set TEMPSORTED;

array DEGREES_F_ {12} 8. (12*.);

DEGREES_F_1=lag(DEGREES_F);

do _i=dim(DEGREES_F_) to 2 by -1;

DEGREES_F_[_i]=DEGREES_F_[_i-1];

end;

run;

Re: Error "Expecting a Name"

Patrick — Sat, 10 Jan 2015 01:50:49 GMT

For the approach you've taken you could simplify your code and reduce volumes and passes through the data as done below:

/* SORT THE DATA BY THE FAKEGROUP AND OBS DATE */

PROC SORT DATA=TEMP OUT=TEMPSORTED;

BY OBS_DATE;

RUN;

/* MAKE A FAKE COPY OF THE DATA AND APPEND TOGETHER INTO 1 DATASET */

DATA TEMPSORTED;

SET TEMPSORTED end=last;

FAKEGROUP=0;

output;

if last then

do;

FAKEGROUP=1;

output;

OBS_DATE+1;

output;

end;

RUN;

/* USE PROC PANEL TO GENERATE THE LAG VARIABLES */

PROC PANEL DATA=TEMPSORTED;

ID FAKEGROUP OBS_DATE;

LAG DEGREES_F(1 2 3 4 5 6 7 8 9 10 11 12) / OUT=TEMP_W_LAG;

RUN;

data TEMP_W_LAG;

set TEMP_W_LAG(where=(FAKEGROUP=0));

run;