Hi: I fear there are too many moving parts and too many unknowns to just guess at what might be wrong, between using TAGSETS.RTF and a custom style template and code that has been tweaked several times since the original code was posted and no data to use, my best recommendation is to open a track with Tech Support, and supply them will ALL the code and enough or a subset or some fake data to reliably replicate the problem. The purpose of TAGSETS.RTF is to allow SAS to control the page breaking based on system options for margins and ODS options or template settings for font size and table spacing. FRAME=HSIDES should be reliable but do you have it only in the template or do you override the style with code in PROC REPORT? That might be part of the factor. the original code used options like headline, headskip, flow and widths specified as percents and a custom style and style overrides for borders and no visible options for the margin settings. This is the type of problem that needs to start from "bare bones" using the default style with TAGSETS.RTF and then slowly, very slowly, add one change at a time until something breaks. That kind of methodical debugging should show you which interaction is causing the output to break as you describe. Cynthia ... View more

Hi:   One suggestion that's been posted is to use BY group processing and the NEWFILE= option. Without more context for your question, it is hard to guess what you mean by different data to different PDF files without doing the same step multiple times. Does that mean something like you want some output to go to 2 PDF files, but other output to only go to one of the 2 PDF files?? It doesn't matter whether you are using PROC REPORT, PROC PRINT, PROC MEANS or PROC REG. Your ODS destination statements go above and below all the procedure code (I call it the "ODS Sandwich").   When a procedure creates output, it writes to ALL the open ODS destinations. Typically, you have only 1 destination open at a time. Sometimes, you have different destinations, like HTML, RTF and PDF all open at the same time. However, if you need to have multiple instances of the SAME destination open to receive output, then one thing you can do is to use the ID= suboption in order to control opening and closing the open destinations. Here's an example using ODS PDF. Notice that I can use either ID= with an identifier or I can just use the identifier: Here's the output (I used STARTPAGE=NO just so all the output would fit on one page for posting): FIRST.PDF   SECOND.PDF   THIRD.PDF   Another way to control what gets written to multiple ODS destinations or the same destination is to use a SAS Macro program to control the output being written. A third way is to write all the output to an ODS DOCUMENT store and then replay selected output objects to the desired destinations. But again, without more context or concrete examples, it is hard to guess.   Cynthia   ... View more

Hi: Those are the instructions for Coursera students. Depending on how the original person is using the Programming 1 class (on our e-learning platform or on the Coursera platform, there ARE different instructions floating out there. Somehow, from the snippet of instructions, I wonder whether the original person is using an instructor's copy of our instructions or possibly following the classroom instructions. The best thing for the original poster to do is to send mail to elearn@sas.com and the instructors on duty can check the elearning roster to see exactly which instructions the person should be following. Usually, for SAS OnDemand for Academics, when the person is working on our elearning server instructions, the setup goes like this: https://youtu.be/piTtnbzODPs -- also our version of the e-learning class has a Virtual Lab with the data already set up, so there is nothing for the student to do, that video is here: https://youtu.be/fs0WpVZnVLM (but this option would not be available to Coursera students). And if the student is using a University-licensed version of SAS to do the work, then there are separate instructions for that data creation. So it is really going to make a difference how this person is taking Programming 1 and which interface they are using. Cynthia ... View more

Hi: Regarding your questions: 1) The default style for ODS EXCEL is typically the HTMLBLUE style. If you want no style, then in the ODS EXCEL statement, after the FILE= option, you would specify STYLE=MINIMAL -- that gives you basically black and white font with no colors. 2) PROC REPORT would support a FORMAT statement like this: format _numeric_ best12.2 _character_ $25.; without needing macro variables or something like this with either 1 macro variable or multiple macro variables: format &list1 dollar14. &list2 comma6. &list3 $25.; Otherwise, you can list multiple variables on a DEFINE statement but for your stated purpose of changing the formats for a group of variables, then the FORMAT statement method would serve that purpose better. Cynthia ... View more

Hi: Without code and data, it's hard to provide any constructive suggestions. Your current output looks like PROC TABULATE output. What PROC REPORT code have you tried? It's not clear to me from your screen shot, which 2 variables need to be divided and whether the division results go in the row (at the bottom of the report) or in a column (to the right side of the report). Can you provide any sample data in the form of a DATA step program? To accomplish crosstabs with PROC REPORT, you have to define ACROSS variable(s). Here's a simple example using SASHELP.CLASS. Getting the sum of the HEIGHT variable is a bit silly, but is shows how to ask for the SUM statistic with a numeric variable under the ACROSS item and also shows the use of the absolute column number in the division:   Without data, it's hard to do more than provide you with this simple example. Cynthia ... View more

Hi: PROC REPORT and PROC TABULATE will generate a lot of reports exactly the same. However, one thing that PROC REPORT has that TABULATE does NOT is the ability to produce customized break lines and add text at the break. One of the things that TABULATE can do that REPORT does NOT do is stack variables in the row dimension of the output table as you show. PROC REPORT would do that in 2 separate report. Essentially TABULATE is generating 2 different tables, the table outlined in black, that you've called out and the table underneath it. These 2 tables are stacked on top of each other in the row dimension using the same variable in the column dimension. PROC REPORT doesn't build tables that way. PROC REPORT builds the report table using the variables in the COLUMN statement and once you define the variables to go across the report row from left to right, you will not be able to change the usage or the nesting of the variables on the report. Nor will you be able to generate 2 separate report tables and stack them. When you need to essentially "stack" tables, such as for a demographic report, you generally need to pre-process the and pre-summarize the data so it is in the structure you want before you create the report output with PROC REPORT. Cynthia ... View more

Hi: This is valid PROC CONTENTS syntax: proc contents data="/home/<userID>/EPG1V2/data/storm_summary.sas7bdat"; run; That is what the first slide is showing. In that example, we are showing the students that it is POSSIBLE to use a fully qualified file name and file extension in the DATA= option of PROC CONTENTS. Then, we show them the LIBNAME syntax. The LIBNAME syntax on SAS OnDemand for Academics for the Programming 1 class would be: libname PG1 "/home/<userID>/EPG1V2/data"; Note that there is NOT a physical filename and file extension on the LIBNAME statement because the LIBNAME statement is pointing to a collection of SAS files, whereas the PROC CONTENTS example with a physical file name was pointing to just one file. So the equivalent PROC CONTENTS for the storm_summary SAS dataset using the 2 level reference would be: proc contents data=PG1.storm_summary; run; Note that when you use the 2 level name -- libname.datasetname -- there are NOT any quotes. However, when you use the physical path location for PROC CONTENTS, there do need to be quotes. I think that somehow the confusion came in understanding the difference between a physical file name shown once in the class and having just one practice in the class. After we show the use of the LIBNAME statement in class, we NEVER show the physical file method again, ever. So that one example was mean to be something of a motivation for why the LIBNAME statement is very useful -- because is makes it possible for you to run the SAME program on 2 different operating systems, by just changing the LIBNAME statment. For example, I can write and test THIS code on my instructor machine: libname PG1 's:\workshop\EPG1V2\data'; proc contents data=PG1.storm_summary; run; proc print data=PG1.storm_summary; run; and then I can give that code to my co-worker on a UNIX system and then ONLY need to change the LIBNAME statement to the location for the data on their system LIBNAME PG1 '/usr/sasdqd/myclass/EPG1V2/data'; ...and ALL the rest of the code could run unchanged. Note that there is NOT any file name or file extension in the LIBNAME statement. Cynthia ... View more

Hi:   I'm not exactly sure what you're asking. If you want to turn off the summary for 2 of the columns, then you need to change the USAGE to DISPLAY, as shown below:     If you only want to add up 2 or 3 of the categories, then you'd need to do some pre-processing or pre-summarizing, probably make some helper variables.  I would make a copy of the dataset like this and turn one category's helper variables into missing: So then, when I do the PROC REPORT, I can move the helper variables under the correct column in the COMPUTE AFTER block: The above screen shot shows the progression from creating the helper variables and using them in 3a) to moving the helper variable values on the summary line over to the report columns (3b) to the final report with the helper variables hidden (3c).   Here's the code for 3c:   cynthia ... View more

Hi: @Tom, that's correct. On the SAS OnDemand for Academics server, the ~ symbol stands as a placeholder for the student's home folder, which is typically located at /home/<userID> so if the userID is u12345678, then these are equivalent LIBNAME statements: libname PG1 '~/EPG1V2/data'; libname PG1 '/home/u12345678/EPG1V2/data'; The reference to ~EPG1V2/data would be incorrect because there needs to be a slash before the EPG1V2 folder name. Cynthia ... View more

Hi: I'm not sure I understand your question. However, the concept you describe "moving proc contents into created Libname folder" indicates that you might not be clear that PROC CONTENTS is showing you information about either all the datasets in a SAS library (defined with a LIBNAME statement) or PROC CONTENTS is showing you information about just 1 dataset in a SAS library (defined with a LIBNAME statement). PROC CONTENTS can point to a library location, but you don't really "move" proc contents into a libname folder. For example, in the Programming 1 class, your data (if you follow the instructions) should be in a class folder labelled EPG1V2 and then the data is in a subfolder named data. So, for example, these are various LIBNAME statements pointing to physical folder locations on different operating systems: libname PG1 'c:\SAS_Class\EPG1V2\data'; libname PG1 '/home/u12345678/EPG1V2/data'; libname PG1 '/usr/bin/sasczz/EPG1V2/data'; libname PG1 's:\workshop\EPG1V2\data'; Then given any of those LIBNAME statements, I could do this: proc contents data=PG1._all_ nods; run; OR proc contents data=PG1.np_summary; run; In one of the first practices in the class, we ask you to use PROC CONTENTS on the storm_summary.sas7bdat file in the class data folder because we want students to understand that they CAN use a physical folder location with PROC CONTENTS. That practice is meant to run like this: proc contents data='<your path to main folder>/EPG1V2/data/storm_summary.sas7bdat'; run; So if you were running on SAS OnDemand for Academics, for example, that code, with the full file path,  would be something like: proc contents data='/home/u12345678/EPG1V2/data/storm_summary.sas7bdat'; run; Then later in the class, we show you how to use a LIBNAME statement to point to the data subfolder for class so that you do not have to use the full physical path in the DATA= option. We also show you in the Programming 1 class, how to use the XLSX Libname engine to see what tables (sheets) are in an Excel workbook. But, none of these examples would be considered moving PROC CONTENTS into a created Libname folder. So if you can provide some more background or clarify your question, then perhaps we can provide more commentary. Cynthia ... View more