About Cynthia_sas

Cynthia_sas

Hi: Without using a LINE statement, you can post-process your data to add a Helper variable called rowtype and then you can assign your pseudo Header values to a new row that you will insert into the data rows, as shown below: Then, your PROC REPORT code will be simplified because now when you run your PROC REPORT on the data file with the helper variable, since you have the pseudo Headers inside the data, you can now use a COMPUTE block and CALL DEFINE to alter the style of each pseudo Header row: I like this method better than the LINE statement method because then I don't have to fiddle with the placement of the text strings on the LINE statement. I did need to add the LENGTH statement to my post-processing step to be sure that the new rows would have a column width big enough for my largest text string. Cynthia

Cynthia_sas

Hi: I'm not sure how you derived your estimate, but the SAS log shows you the exact answer. It looks like you're doing the Practice Exam for the Base Programming Using SAS 9.4 Performance-Based Exam. The datasets you are using are CERT.INPUT08a and CERT.INPUT08b. If you run a PROC CONTENTS on each file BEFORE you run your code, you should see THIS: Assuming that there's a one-to-one match for most of the rows based on ID, just looking at those numbers, I would expect to see at least 2 non-matches. The question asked you to merge the 2 input files BY ID, so without looking at the data, the only way to verify the mismatches and matches is to either manually go through the 2 files row by row or look in the SAS log. Here's the original setup and the 4 questions you have to answer. There are 2 NOTES in the log that answer all 4 questions. I'm not sure why you can't see the SAS Log. All you have to do is run the code. If you are practicing the exam using our copy of the exam, we have a lab where the data is already pre-loaded, so all you have to do is establish your LIBNAME statements to be able to access the data and run code. We use SAS Studio in the Lab for the exam, so you should be able to see the log for the code and for the final PROC CONTENTS. Keep in mind that when you answer the questions, you've got to pay attention to which results file is being asked about MATCH08 or NOMATCH08 and you've got to pay attention whether you're asked to provide the number of observations OR rows; or whether you're asked to provide the number of variables OR columns. My guess is that you're expecting to see 117 variables in both RESULTS.MATCH08 and RESULTS.NOMATCH08; however, remember that you have the DROP= option for RESULTS.NOMATCH08 in the DATA statement for the MERGE. So this means that the ONLY variables left in RESULTS.NOMATCH08 are as shown below: The form of the DROP= option has you excluding ALL the variables that start with "ex" and if you look at the PROC CONTENTS from before you do the merge you'll see that there are a LOT of variables that start with "ex". Cynthia

Cynthia_sas

Hi: I'm confused. I don't see Treatment Name on the report. Which column is the treatment name? You are using so many macro variables it's hard to figure out what your data looks like and which variable is which. My guess is that treatment may be TRTAN or it could be &VARIABLE or confusingly, it could be the column listed as VARIABLE? Just not sure and it looks like you've blanked out most of those column headers anyway. It appears that the leftmost column is just called "VARIABLE" and you've blanked out that header and are using "MRDQ Domain Pearson Correlation (P-value)" as the header over that variable: Is "VARIABLE" the treatment name? If so, I don't understand what you mean when you say you want the label on the same level as treatment name. You're also using custom formats and while some of the custom formats help in figuring out whether variables are character or numeric, it's hard to dummy data without any idea of what the data looks like or what your macro variables and formats are doing. Can you post some dummy data and a version of the code either without macro variables or with macro variable values supplied and either with or without the code for the formats? Cynthia

Cynthia_sas

Hi: I was always taught to add the "ish" to the first color if I wanted to mix colors. However, over the years, I've started not using words for colors and have migrated to using hex color codes which I get from an online HTML color picker. Here's a modified version of your program, where you see that sometimes swapping the color names work (like Red Yellow/Yellow Red), but sometimes it doesn't work like with Blue Green and Green Blue: So I tried Greenish Blue next, but... Honestly, Greenish Blue doesn't look all that "greenish" to me, so when I hear "green blue" I think teal and the RGB hex code for teal is 008080, and that RGB hex value does look like teal to me in the output: If you look at the RGB codes that were used in the ODS HTML output file, you can see exactly what codes were used for each colorname. So, although I can't explain why you get #FF0000 when you specify Green Blue as the color, the workaround is to bite the bullet and find the exact hex code for the color you want. This is the web site I use to find HTML RGB color values: https://www.w3schools.com/colors/colors_picker.asp and I like it because it shows all the different shades lighter to darker if I want to do a gradient: I hope this helps you come up with the a way to specify colors that works for you. Cynthia

Cynthia_sas · ‎11-18-2024

Hi: And, for a concrete example, this paper https://support.sas.com/resources/papers/proceedings13/120-2013.pdf walks you through the process of starting with a working program and eventually converting it into a macro program. I think it illustrates the process that we are all recommending you follow. Cynthia

Cynthia_sas · ‎11-15-2024

Hi: An easier way to "touch" all the fonts used by a changed style template is to just change the font table, as shown below: ods path work.tmp(update) sasuser.templat(update) sashelp.tmplmst(read); proc template; define style Styles.times; parent = styles.pearl; class fonts / 'TitleFont2' = ("Times New Roman, Times",10pt,bold) 'TitleFont' = ("Times New Roman, Times",11pt,bold) 'StrongFont' = ("Times New Roman, Times",8pt,bold) 'EmphasisFont' = ("Times New Roman, Times",8pt,italic) 'headingEmphasisFont' = ("Times New Roman, Times",9pt,bold italic) 'headingFont' = ("Times New Roman, Times",8pt,bold) 'docFont' = ("Times New Roman, Times",8pt); end; run; ods pdf file='c:\temp\times.pdf' style=styles.times; proc print data=sashelp.class; title 'custom style change fonts'; run; proc freq data=sashelp.class; tables age; run; ods pdf close; When I do this, I get the correct borders and the font seems to be used. I only changed the proportional spaced fonts in the font table. I did not bother to change any of the "fixed" or "batch" monospace fonts in the style template definition. Cynthia

Cynthia_sas · ‎11-14-2024

Hi: I assume that you want to see this with a slight break between the header for Asia and the one for Europe. In this example, I did NOT use your template. Instead I just used the simple journal style and made some slight modifications to your REPORT code. First, to avoid warning messages and notes in the log, I changed all of your GROUP usage items to ORDER usage. Next, I did not see the purpose for the X variable, so I didn't use it at all. However, to insert the slight space, I just created a helper variable called BLANKVAR that was computed. Journal style turns off all interior table lines by default, so by inserting the new helper BLANKVAR between Asia and Europe columns, it makes the report look like there's a break and there is, you just need to manually control it. Since you have width=100%, I did not see the need to overcontrol the column widths except that DRIVETRAIN needed to be a bit wider, so I left that. And, then I had to use BLANKVAR in the COLUMN statement. The biggest change I made was in the placement of your border overrides. If you want to alter the borders of the spanning header text strings so they are looking like an underline, then your ODS ESCAPECHAR override needs to appear BEFORE the string, not after as you originally showed. Here's the changed code: I'm also pasting the code below, with an example of how Journal style looks using TEXTDECORATION. But I wanted to provide you with the annotated changes too. Cynthia ** did not change data creation steps although X variable does not seem to be needed; options missing=' ' orientation=portrait topmargin=0.75in bottommargin=0.75in leftmargin=0.5in rightmargin=0.5in; ods escapechar='^'; ods rtf file="c:\temp\span_journal.rtf" style=journal; title1 'Western Region Summary (journal style use borderchanges)'; proc report data=cars2 nowd style(report)={width=100%} style(header)={font_style=roman font_weight=bold}; column type model drivetrain ('^S={borderbottomwidth=1 borderbottomcolor=black}Asia' (Lexus Toyota)) blankvar ('^S={borderbottomwidth=1 borderbottomcolor=black}Europe' (Audi BMW)) ; define Type / order; define drivetrain / "Drive Train" order style(column)={width=30mm}; define model / "Model" order ; define Lexus / display ; define Toyota / display ; define blankvar / computed ' '; define Audi / display ; define bmw / display ; compute blankvar/character length=15; blankvar = '^{nbspace 3} '; endcomp; run; title1 'Western Region Summary (journal style show textdecoration)'; proc report data=cars2 nowd style(report)={width=100%} style(header)={textdecoration=underline font_style=roman font_weight=bold}; column type model drivetrain ("Asia" (Lexus Toyota)) ("Europe" (Audi BMW)) ; define Type / order; define drivetrain / "Drive Train" order style(column)={width=30mm}; define model / "Model" order ; define Lexus / display ; define Toyota / display ; define Audi / display ; define bmw / display ; run; ods rtf close;

Cynthia_sas · ‎11-13-2024

Hi: To expand on @Tom 's helpful reply, here's a way to visualize how dates work in SAS data: A date value is stored internally in a SAS data set as the number of days from the 0 date of Jan 1, 1960. Dates can be both positive and negative numbers when stored. Dates are numeric variables in SAS and ALWAYS stored internally as a number. A negative number means that the date is BEFORE Jan 1, 1960 and a positive number means that the date is AFTER Jan 1, 1960 as represented in the picture above. So -1 would be Dec 31, 1959 and -5 would be Dec 27, 1959. -3334 would be Nov 15, 1950 and -67091 would be Jul 4, 1776. Then positive numbers are the days after Jan 1, 1960. So the number 1 would represent Jan 2, 1960, the number 5 is the stored value for Jan 6, 1960 and the number 8 would be for Jan 9, 1960. Today, Nov 13, 2024, is stored as the number 23693 and Nov 15, 2024 would be stored as 23695. If you have a date value stored as a character string, then you cannot use the character value to calculate values like age or duration. With the date value stored as a number, then simple subtraction will allow you to calculate age in days or duration in days. There are SAS functions that are designed to work specifically with SAS numeric dates, in order to advance dates to a certain point in time or calculate intervals. However, enough said about functions because those are easily found by checking out the documentation on SAS functions. If you are testing a numeric date value in a conditional IF or WHERE statement, then you usually use a date constant in the IF or WHERE statement because, although you could use the internally stored number for date, the date constant makes the program easier to read and maintain if you use a date constant so the condition is understandable. Hope this helps, Cynthia

Cynthia_sas · ‎11-08-2024

Hi, Lisa: The original poster said he could not post his code or data. So here's a simple example using SASHELP.CLASS Note how the call define based on NAME is changing the color of font on the whole row for Alfred to purple. But the tests for age still impact the format for the HEIGHT and WEIGHT columns. With the PROC REPORT ability to change the style and format of a single column, one of the neat things is that the last column listed on the report row can have a COMPUTE block that "touches" or applies styles and formats with CALL DEFINE to any of the other columns on the report row. There are a few more complicated rules around using COMPUTE blocks and CALL DEFINE, but you would learn about those as you read the PROC REPORT documentation and search here in the forums for previous answers to questions like this. Here's the full code I used to create the above output: proc sort data=sashelp.class out=class; by age; run; ods excel file='c:\temp\diff_formats.xlsx' options(embedded_titles='yes'); proc report data=class; title 'Using Call Define to format rows differently based on a condition'; column age name sex height weight; define age / display; define name / display; define sex / display; define height / display; define weight/display; compute weight; ** in the compute block for weight, you can "touch" all the columns on the report row; if name = 'Alfred' then do; call define(_row_,'style','style={font_weight=bold color=purple}'); end; if age = 12 then do; call define('height','format','z8.2'); call define(_col_,'format','z8.2'); end; else if age=14 then do; call define('height','format','6.2'); call define(_col_,'format','6.2'); end; endcomp; run; ods excel close; title; You should be able to run this code because SASHELP.CLASS should be available in every installation of SAS. Hope this helps, Cynthia

Cynthia_sas · ‎11-07-2024

Hi, seeing your single line of PROC REPORT code without seeing ALL the code and ALL the data makes it hard to make a constructive comment. Off the top of my head, I would make sure that the PAGE_2 variable or an automatic break or BY group is not conflicting. You probably have tried tagsets.rtf. I have heard in the past that post-processing RTF files to combine them together can be difficult. (not a suggestion not to try it just an observation that folks who have tried it and succeeded have remarked how hard it was to work out) My other suggestion would be to open a case with Tech Support where they can look at ALL your code and your data and try to help you figure out where the output is breaking. Cynthia

Cynthia_sas · ‎10-22-2024

Hi: I don't see any PROC REPORT code. This seems to be all PROC TEMPLATE and SGRENDER code. Cynthia

Cynthia_sas · ‎10-16-2024

Hi: I'd go back and look at your code more carefully. If your data is as you describe, then you should not have the INPUT function NOTE that you received. Are you sure your weights are standard numbers? I don't observe the same behavior with a test data table that has a character value for WEIGHTOLD and gets converted to a numeric variable WEIGHTNEW: But you can force the INPUT note if your character string is not composed of standard numeric values. In the example below, the period is the thousands separator and the comma is the decimal separator, so I can see the NOTE in the log with a simple INFORMAT and I get rid of the NOTE using the COMMAX informat. Cynthia

Cynthia_sas · ‎10-16-2024

@Tom: Thank you! That's exactly what I would have suggested (escapechar+newline). The other reason that the PROC SQL was confusing was that there were not any ODS Statements shown in the original code. So having the ODS EXCEL statements visible was/is essential. Cynthia

Cynthia_sas · ‎10-16-2024

Hi: That paragraph mark is being displayed in your Microsoft Word because of the RTF control strings created in the RTF file. You can simply go into Word and look for the Show/Hide marks button on the Word menu to turn off the display of the paragraph symbol. As far as I know, 0D0A is the hex code for carriage return/line feed and is not necessarily the same as the code generated inside the RTF. I think that usually the RTF string for a paragraph is \par or \pard...you'd have to look in the RTF. But I don't usually recommend post-processing the RTF control strings because Word is very touchy if you mangle the RTF and then it won't display the file. Cynthia As an example, here's a simple display in Word of an RTF document. The Hide/Show Marks button was turned on in Word. And that is just a small snippet of the RTF file.

Cynthia_sas · ‎10-16-2024

Hi: It's not clear to me what you're doing or what you want. You don't show any ODS statements. You say that you want to "fill data in" the Excel file but it's not clear to me whether the data is coming from SAS or not. I'm not sure why you need a formula. Also, <line break> is not any ODS method for inserting a line feed or line break. And, you have a DATA step that makes your variable VAR from &string. Why do you need the PROC SQL step? Cynthia

Online Status	Offline
Date Last Visited	Thursday