About Reeza

Reeza · ‎03-05-2024

I would probably list a few of the tests that are options and then explain why I choose the one I did over the others. Also, keep in mind the type of test, ie is it a count/chi square more appropriate or continuous/rate data?

Reeza · ‎03-01-2024

Brute force manual method: proc sql; create table integers (num int); insert into integers values(0); insert into integers values(1); insert into integers values(2); insert into integers values(3); insert into integers values(4); insert into integers values(5); insert into integers values(6); insert into integers values(7); insert into integers values(8); insert into integers values(9); quit; proc sql; create table want as select ones.num + tens.num*10 + hundreds.num*100 as total from integers as ones, integers as tens, integers as hundreds having total < 130 order by 1; quit;

Reeza · ‎03-01-2024

SQL is designed to manage data and manipulate it, typically not to create new rows/records. Even a pivot isn't a simple operation in SQL. SAS on the other hand is more data wrangling/analysis friendly. Right tool for the right job. Several different sql options are listed here you can play around with different DBs, and different dbs will have different methods. But not SQL also has the ability to create unique key identifiers on each row automatically, which SAS does not.

Reeza · ‎03-01-2024

data want; set have; by studyID; where firstResult ne 'See Text'; if first.studyID; run; Use FIRST assuming it's sorted and a WHERE to remove the SEE TEXT records. @Gregorytus07 wrote: Hi everyone, I have a dataset that contains multiple result rows for each study ID. I need help in writing a SAS code that selects the first result for each study ID. If the first result is 'See Text', then select the next result value for that study ID. Thanks for your help.

Reeza · ‎03-01-2024

You can get this information, it's just not in one place. There are assumptions of course, that you have labels specified, which are the descriptions and SAS formats are being used. proc sql; create table part1 as select name, label, type, format from sashelp.vcolumn where libname='SASHELP' and memname='CARS'; quit; The second part depends on where you store the formats, assuming you know the formats, you can then do something like the following. proc format cntlout = format_data; select sex_fmt ...; run; That should get you started at least. @zihdonv19 wrote: I expect to get this kind of table (as a separate SAS dataset rather than just display in the "results viewer"): VARIABLE NAME DESCRIPTION/LABEL TYPE/ FORMAT VALUE DESCRIPTION SEX Sex of study participant Char 1 = Male 2 = Female DATE Date of diagnosis Num YYMMDD10. ... ... ...

Reeza · ‎02-28-2024

How do the parameters compare? Not sure if this is the case, but some packages in python standardize the data before doing the regression, SAS does not. Also, what parameterization method is being used if categorical variables are involved. There's likely a reason this is occurring but there isn't enough information to say why. Just running 'logistic regression' in each application is not necessarily running equivalent models. @lionking19063 wrote: Hi, I am constructing a binary logistic regression model using 'proc logistic' in SAS, and the output appears to be error-free with no warning messages. However, when I replicate the same dataset and variables in Python using statsmodels, I encounter an error message indicating possible quasi-complete separation: 'A fraction of 0.20 of observations can be perfectly predicted. This might indicate quasi-separation, and in such cases, some parameters may not be identified.' In light of this discrepancy, I am uncertain about which set of results to trust—those from SAS or from Python. Thank you for any guidance you can provide.

Reeza · ‎02-28-2024

Run the macro with the macro debugging options and check your log. options mprint mlogic; Bad request sounds like a response from the Google site, ie a bad URL request was passed to the filename function.

Reeza · ‎02-28-2024

To expand on my previous response, I would recommend: Reading each file every day and storing as rawData_20240130 (Change date format to YYMMDD for easier usage) Once the files are read into a SAS library and stored correctly then have a second process that uses sashelp.vtable to find the last 6/7 tables. If the dates are in YYMMDD, this is via a sort. proc sql; select catx(".", libname, memname) into :table_list separated by " " from sashelp.vtable where libname='MYLIB' order by memname desc; quit; %put &table_list; Then append all the data together and add the date field in data combined; set &table_list insdsname=src; date=scan(src, -1, "_"); run; Then sort the file and take the latest/earliest record using first/last (I'll assume you know how to do this step)

Reeza · ‎02-28-2024

@User_2024 wrote: Thanks for replying, my issue is not comparing today's output(excel) with yesterday's. i want to compare today's with yesterdays and all of the 6 days prior so I dont have duplicates from any of the files. Currently, my code looks at yesterdays using proc import and eliminates duplicates BUT it is bringing day before yesterday's output to today's. I am so stuck. . my output files are saved as below . for todays run the file saved in yesterdays date. and so on. Before it spits the output for today, it imports the previous day file (26feb) checks for duplicates, avoids it. Now, I want to do this for ALL of the previous files so I dont have ANY duplicates from any of the prev files. Hope this helps. I really appreciate all of your time and help. output_27feb2024 output_26feb2024 output_25feb2024 output_22feb2024 output_21feb2024 Import all the files and stack them using INDSNAME option on the SET statement to add in the filename and extract the date in a new variable. Then sort by ID and descending date and take the first record per ID, then you'll have the latest record for each ID across the 6 files.

Reeza · ‎02-28-2024

Rather than determine file names is it possible to scan the folder and get the latest two file dates?

Reeza · ‎02-28-2024

@sas21 wrote: PROC COMPAREBYIDVARWITHWITH Compare variables of different names But, according to this example, it says the variables need to have different names. And as of now, the two datasets I want to compare have the exact same variable names, which is why I thought it would be better to change the variable names in one of them. Or am I misinterpreting something? You can compare two data sets by ID with the same variable names without having to rename.

Reeza · ‎02-27-2024

I think that thread says you need to use papersize instead of pagesize option. Something like this perhaps (untested): options papersize=("15in","24in"); ods pdf file='/home/fkhurshed/demo_delete.pdf'; proc print data=sashelp.cars; run; options papersize=("8.5in","11in"); proc print data=sashelp.class; run; ods pdf close; @BigPete wrote: According to this thread, it seems like the pagesize= option works with a uniform page size in mind for the same PDF. But is there a way to spit out various page sizes for example like one dataset on 8.5xll pages and another dataset onto 8.5x14 pages for the same PDF file?

Reeza · ‎02-27-2024

Here's a quick way using the dictionary tables to get the list of the variable names, and create the new name using the CATT() function. PROC DATASETS will update the data with the new variable names. *create sample data set; data class; set sashelp.class; run; proc sql noprint; select catx("=", name, catt(name, "_NEW")) /*creates oldvariable = newvariable for rename statement*/ into :rename_list separated by " " from sashelp.vcolumn where libname='WORK' /*replace with your library, or WORK*/ and memname='CLASS' /*replace with data set name - note that it must be upper case, same as libname*/ ; quit; *check the variable rename list in the log; %put &rename_list; *rename the variables without rewriting the full data set; proc datasets library=work nodetails nolist; modify class; rename &rename_list; run; quit; *check output; proc print data=class noobs; run; @sas21 wrote: I have a data set "compare_cohort" with the following variables.... ID COUNTRY BRTHDT DTHDTRACE ......etc it has a lot of variables and I want to make every variable name end in _new, so the new variable name would look like; ID_NEW COUNTRY_NEW BRTHDT_NEW DTHDT_NEW RACE_NEW ETC.... Everything I am trying is adding it to the variable values and not the variable names themselves. Any help is appreciated! Thank you so much!

Reeza · ‎02-26-2024

Because it does 🙂 (Different output options can generate differently styled output). Use the STACKODS option and ODS OUTPUT statement to get a table that looks like the output though. https://documentation.sas.com/doc/en/pgmsascdc/9.4_3.5/proc/p17h6q7ygvkl1sn13qzf947dundi.htm @wwwckk wrote: /* FOLLOWING ARE MY CODES */ title "Base salary discrepancy within ranks"; title2 "By Gender"; PROC MEANS DATA=MD_Data mean; VAR 'Base Salary'n; CLASS 'Department Name'n Grade Gender; output out = PayGap; RUN; The new dataset 'PayGap' looks different from the PROC MEANS table I generated. Thank you!

Reeza · ‎02-26-2024

Doable, yes, efficient, probably not. One option: combine all into one table with the source table identified using INDSNAME option. Then merge for desired results on the two key variables. You could limit to just key values needed if size might be an issue. data combined; set abc.202201-abc.202203 indsname =src; tableName=src; keep key NeededField tableName; run; proc sql; create table want as select a.*, b.needed_field from have as a left join combined as b on upcase(a.TableName)=upcase(b.TableName) order by 1; quit; @ChickenLittle wrote: I have a table that contains a field named TableName, which stores the name of the specific table I need to pull from for each observation (re: each observation requires me to hit against different tables to grab NeededField). For each observation, is there a way to code it so that depending on the table name in the TableName field, I can pull the field I need just from that table for that particular observation? Example: libname abc 'path/of/my/file'; OriginalTable obs key TableName 1 cat abc.202201 2 dog abc.202202 3 frog abc.202303 Table at abc.202201 key NeededField cat 1234 dog 12345 frog 123456 Table at abc.202202 key NeededField cat 345 dog 3456 frog 34567 Table at abc.202202 key NeededField cat 5 dog 56 frog 567 I want to see: OriginalTable obs key TableName NeededField 1 cat abc.202201 1234 2 dog abc.202202 3456 3 frog abc.202303 567 Is this at all possible? Thank you in advance.