Temporary Data Set Error

[MisterJenn]

I want to make a temporary dataset with smokers who are overweight and who were in the alendronate treatment group. However I keep receiving an error message. What am I doing wrong? 

libname hw3 '\\apporto.com\dfs\GWU\Users\kennedyhinnant_gwu\Downloads\Week 3';
data fit1temp;
set hw3.fit2;
if upcase(compress(smoke)) = "CURRENTSMOKER" then currentsmoker = 1;
else if upcase(compress(smoke)) = "NONSMOKER" or 
upcase(compress(smoke)) = "FORMERSMOKER" then currentsmoker = 0;
if blvfx = "yes" or blnspfx = "yes" then any_base_frac = 1;
else if blvfx = "no" and blnspfx = "no" then any_base_frac = 0;
if newvfx = "yes" and newnspfx = "yes" then any_new_frac = 1; 
else if newvfx = "no" and newnspfx = "no" then any_new_frac = 0;
if bmi > 25 then bmi2 = "overweight"; 
else if 0 < bmi <= 25 then bmi2 = "normal";
if 0 < bmi <= 18 then bmi3 = "underweight";
else if 18 < bmi <= 25 then bmi3 = "normal"; 
else if bmi > 25 then bmi3 = "overweight";
bmi1_ind = (bmi > 25);
bmi2_ind = (18 < bmi <= 25);
bmi3_ind = (0 < bmi <= 18);
i_tx = (tx = "Alendronate");
diff_bmd = cobmd-blbmd;
age = (dov - dob)/365.25;
run; 
data fit1temp_tx;
set hw3.fit2;
if tx = "Alendronate";
run; 
data fit1temp_placebo;
set hw3.Fit2;
if tx = "Placebo";
run;
proc sgplot data=fit1temp;
vbox cobmd/group=tx; 
run;
proc means data=fit1temp mean median std p25 p75;
class tx;
var cobmd;
run;
proc ttest data=fit1temp;
class tx;
var cobmd;
run;
data fit1temp_smoke;
set hw3.fit2;
if upcase(compress(smoke)) = "CURRENTSMOKER" then currentsmoker = 1;
else if upcase(compress(smoke)) = "NONSMOKER" or 
upcase(compress(smoke)) = "FORMERSMOKER" then currentsmoker = 0;
if currentsmoker = 1;
run;
proc sgplot data = fit1temp_smoke;
vbox cobmd/group=tx;
run;
proc means data=fit1temp_smoke mean median std p25 p75;
class tx;
var cobmd;
run;
proc npar1way data=fit1temp_smoke wilcoxon; 
class tx;
var cobmd;
run; 
data fit1temp_smoke_ovwt;
set hw3.fit2;
if upcase(compress(smoke)) = "CURRENTSMOKER" then currentsmoker = 1;
else if upcase(compress(smoke)) = "NONSMOKER" or 
upcase(compress(smoke)) = "FORMERSMOKER" then currentsmoker = 0;
if bmi > 25 then bmi2 = "overweight"; 
else if 0 < bmi <= 25 then bmi2 = "normal";
if 0 < bmi <= 18 then bmi3 = "underweight";
else if 18 < bmi <= 25 then bmi3 = "normal"; 
else if bmi > 25 then bmi3 = "overweight";
if currentsmoker = 1; 
if bmi3 = overweight;
diff_bmd = cobmd-blbmd;
proc ttest data fit1temp_smoke_ovwt;
var diff_bmd;
run;

NOTE: Libref HW3 was successfully assigned as follows:
      Engine:        V9
      Physical Name: \\apporto.com\dfs\GWU\Users\kennedyhinnant_gwu\Downloads\Week 3
186  data fit1temp;
187  set hw3.fit2;
188  if upcase(compress(smoke)) = "CURRENTSMOKER" then currentsmoker = 1;
189  else if upcase(compress(smoke)) = "NONSMOKER" or
190  upcase(compress(smoke)) = "FORMERSMOKER" then currentsmoker = 0;
191  if blvfx = "yes" or blnspfx = "yes" then any_base_frac = 1;
192  else if blvfx = "no" and blnspfx = "no" then any_base_frac = 0;
193  if newvfx = "yes" and newnspfx = "yes" then any_new_frac = 1;
194  else if newvfx = "no" and newnspfx = "no" then any_new_frac = 0;
195  if bmi > 25 then bmi2 = "overweight";
196  else if 0 < bmi <= 25 then bmi2 = "normal";
197  if 0 < bmi <= 18 then bmi3 = "underweight";
198  else if 18 < bmi <= 25 then bmi3 = "normal";
199  else if bmi > 25 then bmi3 = "overweight";
200  bmi1_ind = (bmi > 25);
201  bmi2_ind = (18 < bmi <= 25);
202  bmi3_ind = (0 < bmi <= 18);
203  i_tx = (tx = "Alendronate");
204  diff_bmd = cobmd-blbmd;
205  age = (dov - dob)/365.25;
206  run;

NOTE: Missing values were generated as a result of performing an operation on missing values.
      Each place is given by: (Number of times) at (Line):(Column).
      189 at 204:17
NOTE: There were 5580 observations read from the data set HW3.FIT2.
NOTE: The data set WORK.FIT1TEMP has 5580 observations and 22 variables.
NOTE: DATA statement used (Total process time):
      real time           0.04 seconds
      cpu time            0.01 seconds


207  data fit1temp_tx;
208  set hw3.fit2;
209  if tx = "Alendronate";
210  run;

NOTE: There were 5580 observations read from the data set HW3.FIT2.
NOTE: The data set WORK.FIT1TEMP_TX has 2678 observations and 11 variables.
NOTE: DATA statement used (Total process time):
      real time           0.03 seconds
      cpu time            0.00 seconds


211  data fit1temp_placebo;
212  set hw3.Fit2;
213  if tx = "Placebo";
214  run;

NOTE: There were 5580 observations read from the data set HW3.FIT2.
NOTE: The data set WORK.FIT1TEMP_PLACEBO has 2902 observations and 11 variables.
NOTE: DATA statement used (Total process time):
      real time           0.01 seconds
      cpu time            0.01 seconds


215  proc sgplot data=fit1temp;
216  vbox cobmd/group=tx;
217  run;

NOTE: PROCEDURE SGPLOT used (Total process time):
      real time           0.20 seconds
      cpu time            0.07 seconds

NOTE: There were 5580 observations read from the data set WORK.FIT1TEMP.

218  proc means data=fit1temp mean median std p25 p75;
219  class tx;
220  var cobmd;
221  run;

NOTE: There were 5580 observations read from the data set WORK.FIT1TEMP.
NOTE: PROCEDURE MEANS used (Total process time):
      real time           0.01 seconds
      cpu time            0.01 seconds


222  proc ttest data=fit1temp;
223  class tx;
224  var cobmd;
225  run;

NOTE: PROCEDURE TTEST used (Total process time):
      real time           1.34 seconds
      cpu time            0.68 seconds


226  data fit1temp_smoke;
227  set hw3.fit2;
228  if upcase(compress(smoke)) = "CURRENTSMOKER" then currentsmoker = 1;
229  else if upcase(compress(smoke)) = "NONSMOKER" or
230  upcase(compress(smoke)) = "FORMERSMOKER" then currentsmoker = 0;
231  if currentsmoker = 1;
232  run;

NOTE: There were 5580 observations read from the data set HW3.FIT2.
NOTE: The data set WORK.FIT1TEMP_SMOKE has 599 observations and 12 variables.
NOTE: DATA statement used (Total process time):
      real time           0.07 seconds
      cpu time            0.00 seconds


233  proc sgplot data = fit1temp_smoke;
234  vbox cobmd/group=tx;
235  run;

NOTE: PROCEDURE SGPLOT used (Total process time):
      real time           0.12 seconds
      cpu time            0.04 seconds

NOTE: There were 599 observations read from the data set WORK.FIT1TEMP_SMOKE.

236  proc means data=fit1temp_smoke mean median std p25 p75;
237  class tx;
238  var cobmd;
239  run;

NOTE: There were 599 observations read from the data set WORK.FIT1TEMP_SMOKE.
NOTE: PROCEDURE MEANS used (Total process time):
      real time           0.04 seconds
      cpu time            0.03 seconds


240  proc npar1way data=fit1temp_smoke wilcoxon;
241  class tx;
242  var cobmd;
243  run;

NOTE: PROCEDURE NPAR1WAY used (Total process time):
      real time           0.29 seconds
      cpu time            0.17 seconds


244  data fit1temp_smoke_ovwt;
245  set hw3.fit2;
246  if upcase(compress(smoke)) = "CURRENTSMOKER" then currentsmoker = 1;
247  else if upcase(compress(smoke)) = "NONSMOKER" or
248  upcase(compress(smoke)) = "FORMERSMOKER" then currentsmoker = 0;
249  if bmi > 25 then bmi2 = "overweight";
250  else if 0 < bmi <= 25 then bmi2 = "normal";
251  if 0 < bmi <= 18 then bmi3 = "underweight";
252  else if 18 < bmi <= 25 then bmi3 = "normal";
253  else if bmi > 25 then bmi3 = "overweight";
254  if currentsmoker = 1;
255  if bmi3 = overweight;
256  diff_bmd = cobmd-blbmd;

NOTE: Variable overweight is uninitialized.
NOTE: There were 5580 observations read from the data set HW3.FIT2.
NOTE: The data set WORK.FIT1TEMP_SMOKE_OVWT has 0 observations and 16 variables.
NOTE: DATA statement used (Total process time):
      real time           0.09 seconds
      cpu time            0.01 seconds


257  proc ttest data fit1temp_smoke_ovwt;
                     -------------------
                     73
ERROR 73-322: Expecting an =.
258  var diff_bmd;
259  run;

NOTE: The SAS System stopped processing this step because of errors.
NOTE: PROCEDURE TTEST used (Total process time):
      real time           0.01 seconds
      cpu time            0.01 seconds

[SASKiwi · Accepted solution]

Looks like you need at least two corrections - see comments:

data fit1temp_smoke_ovwt;
set hw3.fit2;
if upcase(compress(smoke)) = "CURRENTSMOKER" then currentsmoker = 1;
else if upcase(compress(smoke)) = "NONSMOKER" or 
upcase(compress(smoke)) = "FORMERSMOKER" then currentsmoker = 0;
if bmi > 25 then bmi2 = "overweight"; 
else if 0 < bmi <= 25 then bmi2 = "normal";
if 0 < bmi <= 18 then bmi3 = "underweight";
else if 18 < bmi <= 25 then bmi3 = "normal"; 
else if bmi > 25 then bmi3 = "overweight";
if currentsmoker = 1; 
if bmi3 = "overweight"; * Quotes required:
diff_bmd = cobmd-blbmd;
proc ttest data = fit1temp_smoke_ovwt; * Equals sign required;
var diff_bmd;
run; 

[PaigeMiller]

---

@MisterJenn wrote:

I want to make a temporary dataset with smokers who are overweight and who were in the alendronate treatment group. However I keep receiving an error message. What am I doing wrong? 

---

The only error is here:

257  proc ttest data fit1temp_smoke_ovwt;
                     -------------------
                     73
ERROR 73-322: Expecting an =.
258  var diff_bmd;
259  run;

Is that not self-explanatory?

[Reeza]

It's helpful if you continue from your previous post. 

You shouldn't recreate variables multiple times, it's a waste of resources/programming and if you make a mistake you'll have to fix it in multiple places which is often error prone.

I would recommend changing your code as follows:

data fit1temp_smoke_ovwt;
set fit1temp;

where currentsmoker=1 and bmi2='overweight';
run;

I would also recommend changing your previous step:

data fit1temp_smoke;
set hw3.fit2;
if upcase(compress(smoke)) = "CURRENTSMOKER" then currentsmoker = 1;
else if upcase(compress(smoke)) = "NONSMOKER" or 
upcase(compress(smoke)) = "FORMERSMOKER" then currentsmoker = 0;
if currentsmoker = 1;
run;

to:

data fit1temp_smoke;
set fit1temp;
where currentsmoker = 1;
run;

Or note that you don't even need to create temporary data sets you can apply the filter to the procs directly.

title 'Smokers Analysis';
proc sgplot data = fit1temp;
where currentsmoker=1;
vbox cobmd/group=tx;
run;
proc means data=fit1temp mean median std p25 p75;
where currentsmoker=1;
class tx;
var cobmd;
run;
proc npar1way data=fit1temp wilcoxon; 
where currentsmoker=1;
class tx;
var cobmd;
run; 

title 'Smokers & Overweight Analysis';
proc sgplot data = fit1temp;
where currentsmoker=1 and bmi2='overweight';
vbox cobmd/group=tx;
run;
proc means data=fit1temp mean median std p25 p75;
where currentsmoker=1 and bmi2='overweight';
class tx;
var cobmd;
run;
proc npar1way data=fit1temp wilcoxon; 
where currentsmoker=1 and bmi2='overweight';
class tx;
var cobmd;
run; 

[Astounding]

The main thing you are doing wrong is posting the log here, instead of reading it and trying to let it guide you.

Before you get to the error, the log reports two problems.  You have zero observations where CURRENTSMOKER is 1.  Perhaps the value of SMOKE that you look for ("CURRENTSMOKER") is spelled incorrectly.  But only you can investigate that, nobody here can do that for you.

And you try to use a variable named OVERWEIGHT, but your data doesn't contain such a variable.  Again, perhaps it is a matter of correct spelling, but for whatever reason it is not in your data.  Once again, only you can investigate that.

The only error you get is in PROC TTEST, where you omit the equal sign.  The log pretty clearly spells that out and tells you how to correct it.