SASJedi Tracker

Re: DS2 Code / Model Studio to SAS Studio / SAS VIYA 4

SASJedi — Wed, 13 Aug 2025 17:24:11 GMT

The reference to the ASTORE named AD95B1FC9E115196CE460519A2143F2153D057B9 is not in the DS2 code provided. To move things forward, please provide a copy the DS2 code that created the package named score _BJVQ3JPG5FMTVZRVKDECAY2DF.

Re: Find the count using proc tabulate

SASJedi — Tue, 15 Jul 2025 15:56:50 GMT

When I ran your DATA step as-is, I got the following notes repeatably in the log:

NOTE: Invalid data for event_dt in line 83 21-46.
NOTE: Invalid data for year in line 85 1-14.
RULE: ----+----1----+----2----+----3----+----4----+----5----+----6----+----7----+----8----+----9----+----0                          
85  CH5848162.H.02.6 >=40.30-Aug-22.M.Outpatient.2022                                 
    ZO33333330403323233330332476233040477767666703333222222222222222222222222222222222
    NU584816298902E60ED40930D157D229D9F5401495E492022000000000000000000000000000000000

and the HAVE data set was incomplete and incorrect:

Obs	id	race	age_group	event_dt	sex	visit_type	year
1	5334933	W	02.1 <20 2	.	5790999	35-39 4-	.
2	5956547	W	02.3 25-29	.	5917539	25-29 30	.
3	1002701	B	02.3 25-29	.	5825314	>=40 13-	.

I modified the DATA step as follows to get it to read in all of the DATALINES values properly:

data Have;
length id $7 race $1 age_group $10 event_dt 8 sex $1 visit_type $10 year 8 _event_dt $9;
infile datalines dsd dlm='09'x;
input id race age_group  _event_dt sex $ visit_type $ year:4.;
event_dt=input(compress(_event_dt,'-'),date9.) ;
format event_dt date9.;
datalines;
5334933	W	02.1 <20	21-Apr-20	M	Outpatient	2020
5790999	B	02.5 35-39	4-Jan-21	M	Outpatient	2021
5848162	H	02.6 >=40	30-Aug-22	M	Outpatient	2022
5956547	W	02.3 25-29	9-Jul-20	F	Outpatient	2020
5917539	B	02.3 25-29	30-Jul-20	M	Outpatient	2020
4319890	W	02.3 25-29	9-Jun-20	M	Outpatient	2020
1002701	B	02.3 25-29	30-Sep-20	F	Outpatient	2020
5825314	Z	02.6 >=40	13-Oct-20	M	Outpatient	2020
7047568	O	02.3 25-29	23-Sep-20	M	Outpatient	2020
7205606	B	02.2 20-24	3-Feb-22	F	Outpatient	2022
;
run;

The WANT dataset produced looked like this:

Obs	id	year	sex	age_group	_TYPE_	_PAGE_	_TABLE_	N
1	1002701	2020			1100	1	1	1
2	4319890	2020			1100	1	1	1
3	5334933	2020			1100	1	1	1
4	5790999	2021			1100	1	1	1
5	5825314	2020			1100	1	1	1
6	5848162	2022			1100	1	1	1
7	5917539	2020			1100	1	1	1
8	5956547	2020			1100	1	1	1
9	7047568	2020			1100	1	1	1
10	7205606	2022			1100	1	1	1
11	1002701	.			1000	1	1	1
12	4319890	.			1000	1	1	1
13	5334933	.			1000	1	1	1
14	5790999	.			1000	1	1	1
15	5825314	.			1000	1	1	1
16	5848162	.			1000	1	1	1
17	5917539	.			1000	1	1	1
18	5956547	.			1000	1	1	1
19	7047568	.			1000	1	1	1
20	7205606	.			1000	1	1	1
21		2020	Female		0110	1	1	2
22		2022	Female		0110	1	1	1
23		2020	Male		0110	1	1	5
24		2021	Male		0110	1	1	1
25		2022	Male		0110	1	1	1
26		.	Female		0010	1	1	3
27		.	Male		0010	1	1	7
28		2020		02.1 <20	0101	1	1	1
29		2022		02.2 20-24	0101	1	1	1
30		2020		02.3 25-29	0101	1	1	5
31		2021		02.5 35-39	0101	1	1	1
32		2020		02.6 >=40	0101	1	1	1
33		2022		02.6 >=40	0101	1	1	1
34		.		02.1 <20	0001	1	1	1
35		.		02.2 20-24	0001	1	1	1
36		.		02.3 25-29	0001	1	1	5
37		.		02.5 35-39	0001	1	1	1
38		.		02.6 >=40	0001	1	1	2

I'm not seeing how this output data would be useful in a join with another data set. Can you provide an example of the other dataset you want to join with this data, and a sample of the desired end result?

Re: Proc DS2 - Can my package method OUTPUT to the callers output?

SASJedi — Sun, 22 Jun 2025 15:28:30 GMT

A DS2 method parameter must be one of the supported DS2 data types. DATASET is not a valid datatype. You can use an output statement in the package method to write a row to the dataset that has already been specified in the calling data program.

proc ds2 ;
  package MyComplicatedPackage / overwrite=yes ;
    /* IN_OUT parameters are passed by reference. When this method is called,
       you must specify a variable NAME for the parameter instead of a value.
       The value of the variable becomes accessible to, and can be modified by
	   any of the package methods. */
    method DoComplicatedRecursiveThings (in_out double _z) ;
   /* Change the value to the square of the original */
      _z=_z**2;
   /* Output the result to the calling program's data set */
	   output;
    end ;
  endpackage ;

  data foo / overwrite=yes ;
    declare double zzz ;
    declare package MyComplicatedPackage mcp() ;
    method run () ;
      /* Set the initial value of zzz */
	  zzz=1234;
      /* Modify the value and output using the pacakge method */
      mcp.DoComplicatedRecursiveThings(zzz) ;
    end ;
  enddata ;
run ;
quit;

Results:

Obs	zzz
1	1522756

I hope this helps.

Re: PROC REPORT how to Modify style attribute for a spanning header

SASJedi — Sun, 25 May 2025 20:16:46 GMT

@data_null__ hanks for sharing the answer back here to help future folks find it, too!

Re: Mysterious SAS Message Log Window appears using SAS 9.4 Windows 11

SASJedi — Sun, 04 May 2025 18:18:13 GMT

Man, you just made my day! Thanks 😊

Re: Mysterious SAS Message Log Window appears using SAS 9.4 Windows 11

SASJedi — Mon, 28 Apr 2025 20:40:13 GMT

@PaigeMiller - Grrr... I hate "mysteriously self-healing" computer problems! It always feels like they're just hanging out there waiting to bite you without warning. But I'm glad to hear you're back on track 🙂

Re: Assistance Needed for Upgrading SAS 9.4 M6 to M8 version.

SASJedi — Mon, 28 Apr 2025 19:56:42 GMT

It should be possible to upgrade SAS9 from M6 to M8 directly. For details and requirements, see this (the latest) version of the SAS® 9.4 Guide to Software Updates and Product Changes

Re: Mysterious SAS Message Log Window appears using SAS 9.4 Windows 11

SASJedi — Mon, 28 Apr 2025 19:52:26 GMT

@PaigeMiller - That IS a bummer. I'm using SAS 9.4 (TS1M8) on Windows 11 Enterprise. I tried to reproduce this behavior using this code, but was not successfull:

proc printto log="sas_log_test.txt";
run;
proc print data=sashelp.cars(obs=5);
run;
proc printto;
run;

I tried using a .log extension for the output file, too, just in case that made a difference.
What maintenance level is your SAS9? Might you have any unusual custom SAS configurations that affect SAS logging?

Re: Join vs Merge 1.2 TB with 110 GB Datasets

SASJedi — Mon, 14 Apr 2025 23:59:50 GMT

FedSQL is computationally multi-threaded, but in base SAS, it uses a single read-write thread. In the situation described, the process is most likely I/O bound, not CPU bound. So I don't think FedSQL (or threaded DS2) would help in this situation.

Re: Join vs Merge 1.2 TB with 110 GB Datasets

SASJedi — Fri, 11 Apr 2025 19:11:26 GMT

PROC SQL is not multi-threaded, so there is no threading advantage over DATA step.
SQL and DATA step produce the same result set in a many-to-one situation
Because of the fundamental differences between SQL and DATA step processing, the SQL join will be significantly more resource intensive

I'd recommend using a DATA step MERGE. You've already sorted and indexed the smaller dataset to conform to the larger one's sort order, so you're ready to roll. I'd expect that to process much faster than the SQL join.

Using PROC HTTP 2 - Accessing Internet Information and Debugging

SASJedi — Thu, 24 Apr 2025 18:49:09 GMT

While cURL is the de facto standard for working with APIs from the command line, we’re SAS programmers – and, of course, we want SAS to do the work for us! In the Using FILENAME URL to Access Internet Information post, we experimented with FILENAME URL, but noted that PROC HTTP was significantly more flexible. Let’s explore replicating the cURL code and FILENAME URL examples using PROC HTTP and debugging unexpected results.

Re: Unlocking the Power of Open File Formats: Freedom, Flexibility, and Fierce File Compression

SASJedi — Mon, 07 Apr 2025 12:05:08 GMT

I've been messing around some with storing my SAS data using the parquet LIBNAME engine. I'm impressed with both the reduction in size on disk and the speed of access. It's awesome!

Re: ODS HTML in SAS Viya not working as expected

SASJedi — Wed, 02 Apr 2025 19:19:10 GMT

Ah. OK. If you download the files and review the HTML, you can see that SAS hard-coded the path in the src=specification for the image. It does the same in Windows SAS. Try moving both files to a new location in Windows and you'll see the same effect. To work around this, you'd have to edit the HTML page to change the hardcoded path to this:
src="./SGPlot1.png"

Then, as long as you place both the HTML file and the .png file in the same folder, it will work as expected.

Re: ODS HTML in SAS Viya not working as expected

SASJedi — Wed, 02 Apr 2025 19:09:33 GMT

What version of SAS Viya are you using? I ran this code in Viya 2025.01

ods html file='print.html' path="&_userhome" gpath="&_userhome";
proc print data=sashelp.class(obs=5);
run;
proc sgplot data=sashelp.class;
scatter x=height y=weight;
run;
ods html close;

and got the expected result:

The code also produced the expected results in Viya for Learners (LTS 2024.09).

Re: Efficiently Adding a Shifted Date Column to a Large SAS Dataset

SASJedi — Thu, 27 Mar 2025 20:02:36 GMT

In addition to the integrity constraints mentioned by @mkeintz, indexes on the table may also require modification. So, a couple of questions:

1. Is your data in a SAS data set, or is it stored in a database table accessed via a SAS/ACCESS LIBNAME engine?
If so, the INTNX function cannot be passed to the database. This means that the entire table would have to be pulled out into SAS, the data modified, then copied back into the database. Depending on bandwidth and server co-location issues, this could really choke your performance.

2. Does the table have indexes or integrity constraints defined?
If so, consider dropping the indexes before updating and re-buliding them after the update is complete.

If this is a SAS dataset in a SAS library, consider using DATA step processing instead. Something like this:

/* Create some data to experiment with */
data mylib.have;
	length deal_id date variable 8;
	format deal_id z4. date variable yymmddd10.;
	drop _:;
	retain deal_id 100;
	do _m=1 to 12 by 2;
		do _d=1 to 31 by 3;
			date=mdy(_m,_d,2024);
			if not _error_ then do;
				variable=date;
				output;
			end;
		end;
	end;
run;

title "MYLIB.HAVE - Before mods";
proc print data=mylib.have(obs=5);run;

/* If you can, create both data sets in the same library: */

/* Add the new column, write out to a new data set */
data mylib.want;
	set mylib.have;
	format shifted_variable yymmddd10.;
	shifted_variable=intnx('month', date, -1, 'E');
run;

proc datasets library=mylib nolist nodetails;
   /* The, delete the original */
	delete have;
run;
   /* And rename the new data set with the original data set's name */
	change want=have;
run; quit;

title "MYLIB.HAVE - After mods";
proc print data=mylib.have(obs=5);run;

MYLIB.HAVE - Before mods

Obs	deal_id	date	variable
1	0100	2024-01-01	2024-01-01
2	0100	2024-01-04	2024-01-04
3	0100	2024-01-07	2024-01-07
4	0100	2024-01-10	2024-01-10
5	0100	2024-01-13	2024-01-13

MYLIB.HAVE - After mods

Obs	deal_id	date	variable	shifted_variable
1	0100	2024-01-01	2024-01-01	2023-12-31
2	0100	2024-01-04	2024-01-04	2023-12-31
3	0100	2024-01-07	2024-01-07	2023-12-31
4	0100	2024-01-10	2024-01-10	2023-12-31
5	0100	2024-01-13	2024-01-13	2023-12-31

If your SAS data set had indexes or integrity constraints, they will have been deleted by the DATA step process. You can rebuild them on the new dataset using the PROC DATASETS INDEX CREATE and IC CREATE statements.

Re: sequentially designed datastep run codes within proc cas do not wait for its predecessor to fini

SASJedi — Mon, 24 Mar 2025 19:55:05 GMT

In your second DATA step, this statement:

		if first.timestamp;

ensures that only the first record for each timestamp group is output. So the output data set from the second data step (casuser.result_groups_temp) will most likely have significantly fewer observations than the input data set (casuser.origin_filtered_by_time).

I emulated your process by loading the sashelp.cars dataset to my casuser caslib, and using it in a simplified form of your program. It all worked exactly as I would have expected...

Re: sequentially designed datastep run codes within proc cas do not wait for its predecessor to fini

SASJedi — Mon, 24 Mar 2025 14:15:15 GMT

In what way is the result different from your expected result?

Re: send email from sas with tables in body and text

SASJedi — Mon, 24 Mar 2025 12:54:34 GMT

Try putting the DATA step inside the ODS block, then use FILE PRINT; to direct the DATA step output text to ODS, something like this:

%macro send_Email_From_SAS;
filename outbox email
               to=("Ron.Einstein@BankLeumi.co.il")
               subject="Data sets check"
               type="text/html"
               from='Ron.Einstein@BankLeumi.co.il'
               encoding="utf-8"
;

ods html body=outbox rs=none;

DATA _null_;
   file PRINT;
   Put "היי,"//
    "Here are data sets list:"//
   "r_r.&Overides_data_set.---טבלה ובה פירוט אובריידים--"//
   "r_r.Batch_CS_Exp_Osek_vatik&ABT_mon."//
   "r_r.Batch_CS_Exp_Prati_vatik&ABT_mon."//
   "r_r.Batch_CS_Overide_Osek_vatik&ABT_mon."//
   "r_r.Batch_CS_Overide_Prati_vatik&ABT_mon."//
   "r_r.sofi_lak_Batch_CS&ABT_mon."//
   "r_r.sofi_TZ_Batch_CS&ABT_mon."////
run;

proc print data= temp noobs;
   by model;
   var  var_name OR_impact;
run;
ods html  close;
%mend send_Email_From_SAS;

%send_Email_From_SAS

Re: creating episodes using multiple dates when not clean sorted

SASJedi — Fri, 21 Mar 2025 12:31:05 GMT

I don't have the bandwidth to work on this right now, but here's code to create the HAVE dataset for testing:

data have;
  infile datalines dsd truncover;
  input ID:32. trgt_dt:32. A1600:32. A1700:32. A1800:32. R4:32. A2100:32.;
  format trgt_dt A1600 R4 DATE11.;
datalines4;
1,14770,,,,,
1,14775,14763,1,3,,
1,14783,,,,14783,1
1,16022,,,,,
1,16028,16015,1,3,,
1,16031,,,,16031,1
1,17189,17182,1,3,,
1,17195,,,,,
1,17197,,,,17197,1
1,17198,,,,,
1,17208,17197,1,3,17208,3
1,17209,17209,2,3,,
1,17216,17203,1,3,,
1,17219,,,,,
1,17230,,,,,
1,17251,,,,17251,1
;;;;

Re: Why do i use the %DO %UNTIL loop with the %SCAN macro function make a report each year in the li

SASJedi — Mon, 10 Mar 2025 17:43:53 GMT

The first thing I see is that your code does not provide a space-delimited list of years for the macro to work with. If you provide some values like this:

%storms(1980 1990 2000 2010)

your macro will generate a report for each year.