SAS Data Step if statement not working - sas

With the below code the email always gets sent. 1 obviously does not equal 0, but yet it still runs. I have tried removing the do part but still get the same issue.
data _null_;
set TestTable;
if 1 = 0 then do;
file sendit email
to=("email#gmail.com")
subject="Some Subject Line";
end;
run;

While the file statement is considered an executable statement (and thus should not be executed when behind a false if statement), that is not really entirely true. SAS sees the file statement during compilation, and knows that it needs to create a file to write to - and thus, it's somewhat compile-time. That's what is happening here - SAS creates the file (in this case, the email) as a result of the compiler's activity, then doesn't actually populate it with anything, but still has an email at the end of the day.
The same thing happens with any other file - like so:
data _null_;
set sashelp.class;
if 0 then do;
file "c:\temp\test_non_zero.txt";
put name $;
end;
run;
A blank file is created by that code.
If you need to conditionally send emails, I would recommend wrapping your email code in a macro, and then calling that macro using call execute or similar from the dataset. Like so:
%macro write_email(parameters);
data _null_;
file sendit email
to=("email#gmail.com")
subject="Some Subject Line";
run;
%mend write_email;
data _null_;
set TestTable;
if 0 then do;
call execute('%write_email(parameters)');
end;
run;

Use email directives to abort the message.
http://support.sas.com/documentation/cdl/en/lrdict/64316/HTML/default/viewer.htm#a002058232.htm
data _null_;
file sendit email to=("email#gmail.com") subject="Some Subject Line";
if nobs=0 then put '!EM_ABORT!';
set TestTable nobs=nobs;
....
run;

Related

SAS Macro variable escaping apostrophe in variable name Proc Http

I have been working on this for 3 days now and have tried all I can think of including %str(),%bquote(), translate() and tranwrd() to replace single apostrophe with double apostrophe or %’
The below data step and macro work fine until I hit a last name which contains an apostrophe e.g. O’Brien. I then encounter syntax errors due to un closed left parentheses. The below code I have left what I thought was closest to working with the tranwrd included.
Any assistance you can provide is greatly appreciated.
%macro put_data (object1,id);
Proc http
method=“put”
url=“https://myurl/submissionid/&id”
in=&object1;
Headers “content-type”=“application/json”;
Run;
%mend;
data _null_;
Set work.emp_lis;
call execute(catt(‘%put_data(‘,’%quote(‘’{“data”:{“employeeName”:”’,tranwrd(employeeName,”’’”,”’”),’”}}’’),’,id,’)’));
run;
Craig
There are a wide potential of problems in constructing or passing a json string in SAS macro. Proc JSON will produce valid json (in a file) from data and that file in turn can be specified as the input to be consumed by your web service.
Example:
data have;
length id 8 name $25;
input id& name&;
datalines;
1 Homer Simpson
2 Ned Flanders
3 Shaun O'Connor
4 Tom&Bob
5 'X Æ A-12'
6 Goofy "McDuck"
;
%macro put_data (data,id);
filename jsonfile temp;
proc json out=jsonfile;
export &data.(where=(id=&id));
write values "data";
write open object;
write values "name" name;
write close;
run;
proc http
method="put"
url="https://myurl/submissionid/&id"
in=jsonfile
;
headers "content-type"="application/json";
run;
%mend;
data _null_;
set have;
call execute(cats('%nrstr(%put_data(have,',id,'))'));
run;
I was able to find issue with my code with the tranwrd statement was backwards and it needed to be moved to the proc sql create table statement. I also needed to wrap &object1 in %bquote. This was the final code that worked.
When creating table wrap variables in tranwrd as below.
tranwrd(employeeName, “‘“,”’’”)
% macro put_data (object1,id);
Proc http
method=“put”
url=“https://myurl/submissionid/&id”
in=%bquote(&object1);
Headers “content-type”=“application/json”;
Run;
%mend;
data _null_;
Set work.emp_lis;
call execute(catt(‘%put_data(‘,’%quote(‘’{“data”:{“employeeName”:”’,employeeName,’”}}’’),’,id,’)’));
run;
Just use actual quotes and you won't have to worry about macro quoting at all.
So if your macro looks like this:
%macro put_data(object1,id);
proc http method="put"
url="https://myurl/submissionid/&id"
in=&object1
;
headers "content-type"="application/json";
run;
%mend;
Then the value of OBJECT1 would usually be a quoted string literal or a fileref. (There are actually other forms.) Looks like you are trying to generate a quoted string. So just use the QUOTE() function.
So if your data looks like:
data emp_lis;
input id employeeName $50.;
cards;
1234 O'Brien
4567 Smith
;
Then you can use a data step like this to generate one macro call for each observation.
data _null_;
set emp_lis;
call execute(cats
('%nrstr(%put_data)('
,quote(cats('{"data":{"employeeName":',quote(trim(employeeName)),'}}'))
,',',id
,')'
));
run;
And your SAS log will look something like:
NOTE: CALL EXECUTE generated line.
1 + %put_data("{""data"":{""employeeName"":""O'Brien""}}",1234)
NOTE: PROCEDURE HTTP used (Total process time):
real time 2.46 seconds
cpu time 0.04 seconds
2 + %put_data("{""data"":{""employeeName"":""Smith""}}",4567)
NOTE: PROCEDURE HTTP used (Total process time):
real time 2.46 seconds
cpu time 0.04 seconds

ERROR: No logical assign for filename FNAME

I was running the following code:
%LET TIME_INTERVAL='MINUTE15';
/*
* Get the file names of a specific location
*/
%MACRO get_filenames(location,filenames);
filename _dir_ "%bquote(&location.)";
data &filenames(keep=fname);
handle=dopen( '_dir_' );
if handle > 0 then do;
count=dnum(handle);
do i=1 to count;
fname=dread(handle,i);
output &filenames;
end;
end;
rc=dclose(handle);
run;
filename _dir_ clear;
%MEND;
%MACRO NBBO (fname);
DATA TICKERS_NBBO;
INFILE &fname;
INPUT SYMBOL $;
RUN;
%mend;
%MACRO CALCU(DATE_VAR);
%get_filenames('./groups',filenames);
data _null_;
set filenames;
by fname;
if fname =: "&TIME_INTERVAL";
%NBBO(fname);
run;
%mend;
However, I got the error: ERROR: No logical assign for filename FNAME.
I was wondering what is the reason that caused this?
There are lots of csv files in the folder groups. I was trying to run the NBBO macro on each of the files and load each of the files into a dataset with infile statement.
You're mixing up data step code and macro code in a manner that's not permitted. You can't extract the contents of the variable fname and use it to populate a macro variable in this way. Instead you're passing the text fname to that macro variable, which then doesn't work.
You also can't run another data step inside a data step in this manner. You need to use one of the various methods for executing it after the data step terminates or during execution.
Finally, it's entirely unclear to me what you're really doing here: because you're going to end up with just one of those files in the dataset.
I think you may want something like this, which is much easier. I use the FILEVAR option in infile, which uses the value of a variable; I have to make some changes to how you calculate fname (I think you have to do this anyway, it has no directory name in it).
%MACRO get_filenames(location,filenames);
filename _dir_ "%bquote(&location.)";
data &filenames(keep=fname);
length fname $512;
handle=dopen( '_dir_' );
if handle > 0 then do;
count=dnum(handle);
do i=1 to count;
fname=catx('\',"%bquote(&location.)",dread(handle,i));
output &filenames;
end;
end;
rc=dclose(handle);
run;
filename _dir_ clear;
%MEND;
%get_filenames(location=c:\temp\cars, filenames=fnam);
data tickers_nbbo;
set fnam;
infile a filevar=fname dlm=',';
input symbol $;
run;
If you really do need to call a data step separately, you either need to use CALL EXECUTE, DOSUBL, or construct macro calls in another way (PROC SQL SELECT INTO, %INCLUDE, etc.).

SAS %errmail macro

Although I have the error handling:
filename myfile email
to=&e_mail.
subject= "Error: &number."
type="text/plain";
%macro errmail;
%if &syserr ne 0 %then %do;
options obs= max replace nosyntaxcheck;
data _null_;
file myfile;
put;
put 'ERROR';
put "&syserrortext";
put 'check a log'
run;
%abort cancel;
%end;
%mend errmail;
when I have the error in the proc export:
(&number. is the table in the work)
proc export data=&number.
outfile="/usr/local/backup/&number./&number._%SYSFUNC(TODAY(),DATE9.).txt"
replace
dbms=dlm;
delimiter=';';
run;
%errmail;
ERROR: Physical file does not exist, /usr/local/backup/2116/2116_13MAY2016.txt.
NOTE: The SAS System stopped processing this step because of errors.
NOTE: There were 1 observations read from the data set WORK.2116.
NOTE: DATA statement used (Total process time):
real time 0.01 seconds
cpu time 0.01 seconds
SAS dosen't check a macro and doesn't send an e-mail. I have an error in the path, so I understand the error, but I need an e-mail about that and I want that SAS stops processing because of macro.
Adding some options to the macro can fix it?
this is how I send out my email from SAS
not required to create a file on the system
filename outbox email "email#address.com";
data _null_;
file outbox
to=("email#address.com")
cc=("email#address.com")
subject="SUBJECT OF EMAIL "
/* attach=("C:\sas\results.out" "C:\sas\code.sas") */ ; /* incase you need to attach a file */
put 'Hello! ';
put ' ';
put 'Thank you & let me know if there is anything,';
put "&macrovar."; /*in case you need to put some detail in the email from macro variables make sure to use double quote for that */
put 'SIGNATURE ME';
run;

sas insert table into html email

Is it possible to loop through the records of a table to populate an html email without repeating the beginning and the end of the email?
With this example I get a mail with 5 tables of 1 row (because WORK.MyEmailTable is table of 5 records and set creates a loop in the data step):
data _null_;
file mymail;
set WORK.MyEmailTable;
put '<html><body><table>';
***loop through all records;
put '<tr>';
put %sysfunc(cats('<td>',var1,'</td>'));
put %sysfunc(cats('<td>',var2,'</td>'));
put %sysfunc(cats('<td>',var3,'</td>'));
put '</tr>';
put '</table></body></html>';
run;
And I'm looking to have 1 table of 5 rows.
I don't know if there is a way to prevent recursively put the beginning and the end of the mail when you use set in the data step.
(Let me know if it's not clear I'll update.)
Thank you,
You can use the _n_ automatic datastep variable to let you know when you are on the first observation, and the set statement option end= to know that you are on the last observation:
data _null_;
file mymail;
set WORK.MyEmailTable end=eof;
if _n_ eq 1 then do;
put '<html><body><table>';
end;
/*loop trhough all records*/
put '<tr>';
put %sysfunc(cats('<td>','_n_=',n,' eof=',eof,' ',var1,'</td>'));
put %sysfunc(cats('<td>','_n_=',n,' eof=',eof,' ',var2,'</td>'));
put %sysfunc(cats('<td>','_n_=',n,' eof=',eof,' ',var3,'</td>'));
put '</tr>';
if eof then do;
put '</table></body></html>';
end;
run;
I've added the values _n_ and eof to the output so you can see clearly how they work.
Rob's method is pretty much the standard, but there is another option if you prefer scripting an explicit loop (which can be more comfortable for non-SAS programmers to read). This will function exactly like Rob's answer, and may well compile to the same machine code even.
data _null_;
file mymail;
put '<html><body><table>';
do _n_ = 1 by 1 until (eof);
/*loop trhough all records*/
set WORK.MyEmailTable end=eof;
put '<tr>';
put %sysfunc(cats('<td>',var1,'</td>'));
put %sysfunc(cats('<td>',var2,'</td>'));
put %sysfunc(cats('<td>',var3,'</td>'));
put '</tr>';
end;
put '</table></body></html>';
stop;
run;
_n_ here doesn't have any special meaning (like it does in Rob's answer); it's used by convention since this way it does effectively have the same meaning as it does normally.
You need to use the end=eof to create a variable eof which is true on the last record of the dataset; otherwise the data step will terminate prematurely (before actually hitting your final statement). You also need the stop to tell it to not go back to the start - otherwise it will, and will put a new starting section, then terminate instantly when it hits the set. (Try it and see.)
do _n_=1 by 1 until (eof); is a SAS-specific way of using an incremental loop; it's similar to the c/c++ for (_n_=1; !eof; _n_++) for example - it allows you to have an auto-incremented do loop whilst having a separate, unrelated stopping criteria.

Repeatedly running a data step if URL fails to load

I am creating a dataset using filename URL web submissions. However, in some instaces I keep getting '502' responses from the server. To get around this I would like to use some conditional logic inside a macro. I'm most of the way there, but I cant quite get the end bit to work. The idea is that the macro, which is nested within other nested macros will keep trying this one submission until it gets a dataset that doesn't have 0 observations then move on:
%macro test_exst;
filename loader url "http://finance.yahoo.com/d/quotes.csv?s=&svar1.+&svar2.+&svar3.+&svar4.+&svar5.+&svar6.+&svar7.+&svar8.+&svar9.+&svar10.+
&svar11.+&svar12.+&svar13.+&svar14.+&svar15.+&svar16.+&svar17.+&svar18.+&svar19.+&svar20.+
&svar21.+&svar22.+&svar23.+&svar24.+&svar25.+&svar26.+&svar27.+&svar28.+&svar29.+&svar30.+
&svar31.+&svar32.+&svar33.+&svar34.+&svar35.+&svar36.+&svar37.+&svar38.+&svar39.+&svar40.+
&svar41.+&svar42.+&svar43.+&svar44.+&svar45.+&svar46.+&svar47.+&svar48.+&svar49.+&svar50.+
&svar51.+&svar52.+&svar53.+&svar54.+&svar55.+&svar56.+&svar57.+&svar58.+&svar59.+&svar60.+
&svar61.+&svar62.+&svar63.+&svar64.+&svar65.+&svar66.+&svar67.+&svar68.+&svar69.+&svar70.+
&svar71.+&svar72.+&svar73.+&svar74.+&svar75.+&svar76.+&svar77.+&svar78.+&svar79.+&svar80.+
&svar81.+&svar82.+&svar83.+&svar84.+&svar85.+&svar86.+&svar87.+&svar88.+&svar89.+&svar90.+
&svar91.+&svar92.+&svar93.+&svar94.+&svar95.+&svar96.+&svar97.+&svar98.+&svar99.+&svar100.+
&svar101.+&svar102.+&svar103.+&svar104.+&svar105.+&svar106.+&svar107.+&svar108.+&svar109.+&svar110.+
&svar111.+&svar112.+&svar113.+&svar114.+&svar115.+&svar116.+&svar117.+&svar118.+&svar119.+&svar120.+
&svar121.+&svar122.+&svar123.+&svar124.+&svar125.+&svar126.+&svar127.+&svar128.+&svar129.+&svar130.+
&svar131.+&svar132.+&svar133.+&svar134.+&svar135.+&svar136.+&svar137.+&svar138.+&svar139.+&svar140.+
&svar141.+&svar142.+&svar143.+&svar144.+&svar145.+&svar146.+&svar147.+&svar148.+&svar149.+&svar150.+
&svar151.+&svar152.+&svar153.+&svar154.+&svar155.+&svar156.+&svar157.+&svar158.+&svar159.+&svar160.+
&svar161.+&svar162.+&svar163.+&svar164.+&svar165.+&svar166.+&svar167.+&svar168.+&svar169.+&svar170.+
&svar171.+&svar172.+&svar173.+&svar174.+&svar175.+&svar176.+&svar177.+&svar178.+&svar179.+&svar180.+
&svar181.+&svar182.+&svar183.+&svar184.+&svar185.+&svar186.+&svar187.+&svar188.+&svar189.+&svar190.+
&svar191.+&svar192.+&svar193.+&svar194.+&svar195.+&svar196.+&svar197.+&svar198.+&svar199.+&svar200.
&f=&&fvar&a." DEBUG ;
/* data step based on filename url above goes here, each pass will give 500 metrics x 1 symbol dataset*/
%put create dataset from csv submission;
data temp_&I._&&fvar&a.;
infile loader length=len MISSOVER /*delimiter = ','*/;
/* input record $varying8192. len; */
input record $varying30. len;
format record $30.;
informat record $30.;
run;
data _null_;
dsid=open("temp_&I._&&fvar&a.");
obs=attrn(dsid,"nobs");
put "number of observations = " obs;
if obs = 0 then stop;
else;
filename loader url "http://finance.yahoo.com/d/quotes.csv?s=&svar1.+&svar2.+&svar3.+&svar4.+&svar5.+&svar6.+&svar7.+&svar8.+&svar9.+&svar10.+
&svar11.+&svar12.+&svar13.+&svar14.+&svar15.+&svar16.+&svar17.+&svar18.+&svar19.+&svar20.+
&svar21.+&svar22.+&svar23.+&svar24.+&svar25.+&svar26.+&svar27.+&svar28.+&svar29.+&svar30.+
&svar31.+&svar32.+&svar33.+&svar34.+&svar35.+&svar36.+&svar37.+&svar38.+&svar39.+&svar40.+
&svar41.+&svar42.+&svar43.+&svar44.+&svar45.+&svar46.+&svar47.+&svar48.+&svar49.+&svar50.+
&svar51.+&svar52.+&svar53.+&svar54.+&svar55.+&svar56.+&svar57.+&svar58.+&svar59.+&svar60.+
&svar61.+&svar62.+&svar63.+&svar64.+&svar65.+&svar66.+&svar67.+&svar68.+&svar69.+&svar70.+
&svar71.+&svar72.+&svar73.+&svar74.+&svar75.+&svar76.+&svar77.+&svar78.+&svar79.+&svar80.+
&svar81.+&svar82.+&svar83.+&svar84.+&svar85.+&svar86.+&svar87.+&svar88.+&svar89.+&svar90.+
&svar91.+&svar92.+&svar93.+&svar94.+&svar95.+&svar96.+&svar97.+&svar98.+&svar99.+&svar100.+
&svar101.+&svar102.+&svar103.+&svar104.+&svar105.+&svar106.+&svar107.+&svar108.+&svar109.+&svar110.+
&svar111.+&svar112.+&svar113.+&svar114.+&svar115.+&svar116.+&svar117.+&svar118.+&svar119.+&svar120.+
&svar121.+&svar122.+&svar123.+&svar124.+&svar125.+&svar126.+&svar127.+&svar128.+&svar129.+&svar130.+
&svar131.+&svar132.+&svar133.+&svar134.+&svar135.+&svar136.+&svar137.+&svar138.+&svar139.+&svar140.+
&svar141.+&svar142.+&svar143.+&svar144.+&svar145.+&svar146.+&svar147.+&svar148.+&svar149.+&svar150.+
&svar151.+&svar152.+&svar153.+&svar154.+&svar155.+&svar156.+&svar157.+&svar158.+&svar159.+&svar160.+
&svar161.+&svar162.+&svar163.+&svar164.+&svar165.+&svar166.+&svar167.+&svar168.+&svar169.+&svar170.+
&svar171.+&svar172.+&svar173.+&svar174.+&svar175.+&svar176.+&svar177.+&svar178.+&svar179.+&svar180.+
&svar181.+&svar182.+&svar183.+&svar184.+&svar185.+&svar186.+&svar187.+&svar188.+&svar189.+&svar190.+
&svar191.+&svar192.+&svar193.+&svar194.+&svar195.+&svar196.+&svar197.+&svar198.+&svar199.+&svar200.
&f=&&fvar&a." DEBUG ;
data temp_&I._&&fvar&a.;
infile loader length=len MISSOVER /*delimiter = ','*/;
/* input record $varying8192. len; */
input record $varying30. len;
format record $30.;
informat record $30.;
run;
run;
%mend;
%test_exst;
The idea here is try URL submission, create dataset from it, check number of obs is not zero. If its not end the macro. If it is resubmit the same filename URL then create the dataset from it again. Keep doing this until the server respond, the exit the macro and most on to the rest of the code.
I haven't got as far as running this code in anger yet. I'm guessing the filename URL will work fine, but it is the fact that the code is attempting to create a dataset within a data null step right at the end that is making it fall over. Any ideas?
Thanks
Without getting into the specifics of your project, a good way to approach this generally is with recursion.
%macro test_me(iter);
%let iter=%eval(&iter.+1);
data my_data;
infile myfilename;
input stuff;
call symputx("obscount",_n_);
run;
%if &obscount=1 and &iter. < 10 %then %do;
%put Iteration &iter. failed, trying again;
%test_me(&iter.);
%end;
%mend test_me;
%test_me(0);
It checks to see if it worked, and if it did not work, it calls itself again, with a maximum iteration count to make sure you don't end up in infinite loop land if the server is down or somesuch. You also might put a delay in there if the server has a maximum frequency of calls or any other rules the API requires you to follow.