ERROR: Variable _29MAR2016 not found

[matteofrancioni]

Hi everybody,

I really can't understand why i got this message "ERROR: Variable _29MAR2016 not found" while i'm trying to perfom a PROC GLM on my SAS University Edition. Of course I've tried to change the format of date on excel but dindn't work at all.

I need to perfom ANOVA with repeated measures in a excel databse like this one: 

Date	Treatment	HSR
29/03/2016	A	0,5
29/03/2016	B	0,2
29/03/2016	C	1,1
27/04/2016	A	0,2
27/04/2016	B	0,5
27/04/2016	C	0,9
10/05/2016	A	1,2
10/05/2016	B	1,1
10/05/2016	C	0,5

The script is the following:

PROC PRINT; RUN; data new1; set import;

keep treatment date HSR;

run;

proc sort data=new1;

by  treatment;

run;

proc print;

run;

proc transpose out=New2;

by treatment;

id date;

run;

proc print;

run;

data new3; set new2;

drop _name_ _label_;

run;

proc print; run;

PROC GLM data=new3;

CLASS  treatment;

title "CO2_flux";

model _29MAR2016--_10MAY2016 =  treatment;

repeated time 3 (1 2 3) / summary printe;

MEANS treatment / lsd alpha=0.05; run;

Many thanks to everyone who will try to help me!

Cheers!

[Kurt_Bremser]

The model statement needs variables, not values, as arguments.

[RW9]

Not really familiar with proc glm myself, but my question would be why are you transposing dates up to be columns?  That really doesn't sound like a good apprach.

[ballardw]

Show us the result of proc contents for data=new3.

[matteofrancioni]

Ok, I think I solved this issues using this script:

PROC PRINT DATA=WORK.MYEXCEL; RUN;
PROC PRINT; RUN;
data new1; set import (rename=data=date);
keep date crop plot TX;
run;
proc print; run;

proc sort data=new1;
by crop plot;
run;
proc print; run;

proc transpose out=New2;
by crop plot;
id date;
run;
proc print; run;

data new3; set new2;
drop _name_ _label_;
run;
proc print; run;

PROC CONTENTS DATA=new3; RUN;

PROC GLM data=new3;
CLASS crop ;
title "RH";
model _06JAN2015--_28DEC2015 = crop ;
repeated time 20 (1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20) / summary printe;
MEANS crop / lsd alpha=0.05; run;

On a database that had 20 dates of measurements and it was structured like this (example with just 2 dates)

Date	Plot	Crop	TX
06-Jan-15	13	Wheat	0,1
06-Jan-15	15	Wheat	0,3
06-Jan-15	17	Wheat	0
06-Jan-15	19	Corn	0,1
06-Jan-15	21	Corn	0,65
06-Jan-15	23	Corn	0,95
06-Jan-15	25	Tobacco	0,1
06-Jan-15	27	Tobacco	0,1
06-Jan-15	29	Tobacco	0
03-Apr-15	13	Wheat	6,75
03-Apr-15	15	Wheat	5,8
03-Apr-15	17	Wheat	6,45
03-Apr-15	19	Corn	6,7
03-Apr-15	21	Corn	6,25
03-Apr-15	23	Corn	6,3
03-Apr-15	25	Tobacco	6,15
03-Apr-15	27	Tobacco	5,95
03-Apr-15	29	Tobacco	2,93