I'm just reaching out for some advice.   I am comparing two data tables.  I am reporting the changes to an Excel file with 3 tabs: one for added records, one for deleted records, and one for modified records.  In the modified tab, I would like to highlight just the cells that changed from the previous version to the current version so there isn't necessity for the reviewer to search each variable for the changes.  I can highlight cells based on conditions using the COMPUTE statement in PROC REPORT.  Is there a way to flag the modified variables so I can highlight only changed values?   Below is a small example.  In my output, the Added tab would show Bill, the Deleted tab would show Alice, and the Modified tab would show Jane and Phillip.  The only modification I want to make is that WEIGHT would be highlighted for Jane and AGE would be highlighted for Philip since those are the only change.   Any suggestions on how to do this?   proc sort data=sashelp.class out=old; by name; run; proc sql; create table new as select * from sashelp.class where 1=0; insert into new values ('Bill','M',13,65,82); quit; data new; set old new; by name; if name='Alice' then delete; if name='Philip' then age=14; run; data compared (keep=status name sex age height weight); length STATUS $10; merge new (in=a) old (in=b rename=(sex=sex_old age=age_old height=height_old weight=weight_old)); by name; if b and not a then status='DELETED'; else if a and not b then status='ADDED'; else if a & b then do; status=ifc(sex=sex_old & age=age_old & height=height_old & weight=weight_old, '', 'MODIFIED'); end; run;   ... View more

I have recently changed jobs and now have a new email address that I would like to use for this community.  I cannot find in my account preferences how to change the email address associated with my account.  I am still temporarily consulting for my previous employer, so I am still able to access my old email for a while, but I would like to switch over anyway.   I do not want to have to make a new account because I would like to still have all the questions that I have asked in the past as reference.  Anyone know how to make the change?   ... View more

I have data that I am exporting from SAS to Excel for our data managers.  There are a couple of issues that I am noticing that I cannot correct.  I want to see if there is a way to change this in SAS instead of manually doing it in Excel.   First, we have site numbers that are 3-digit character values, usually containing leading zeros.  For example, site 001, 002, 003, etc.  When it exports into Excel, it displays as numeric 1, 2, 3, etc.  Is there a way to keep the leading zeros in Excel?  If it can be retained as a character value, that would be ideal, but not necessary if we can at least keep the leading zeros.   FYI, I have tried adding to the SAS variable an apostrophe since Excel uses an apostrophe to treat a numeric value as a character string, but that does not work.   Second, I have long ID variables containing values like "5086074278593750", but they display in Excel as "5.08607E+15".  I want to ignore the scientific notation and display it as the raw value.  I can manually change the format in Excel from general to number and then remove the 2 decimal places, but I don't want to have to do that every time to every variable that is reformated.   In short, I want my data to show up in Excel exactly as I have it in SAS.  Not format changes.  Any solutions would be much appreciated! ... View more

I have a list of lab values in a character variable, and I need to store all the numeric values in a numeric variable.  I thought I had it made when I found the NOTDIGIT() function.  That works great for integer data, but then I realized that I am dealing with decimal and negative values.  Is there a way to detect just these characters?  I know I could just use INPUT().  That would convert any number-like value to numeric, but for all the character values, I would be left with a lot of ugly notes filling up my log (NOTE: Invalid argument to function INPUT at line XXX column YYY.).   I have some example data below.  Is there  away to get LABVAL to the NUMLAB and CHARLAB, respectively, as I have them?   LABVAL NUMLAB CHARLAB 7 7   NEG   NEG -3 -3   +   + 2-4   2-4 1.5 1.5   2+   2+ 98.6 98.6   12.773 12.773   YELLOW   YELLOW < 2.5   < 2.5 ... View more

I have a scenario that I have been coming across more and more often.  I figured now is the time to see if there is a more efficient way to do what I want.   Suppose I have 2 tables: an adverse event table and a visit table (see below for examples).  I might want to take the date that an AE occurred (AEDT) and see in which cycle of the study the AE occurred.   SUBJID AEDT TERM 001-001 1/7/2016 Nausea 001-001 1/9/2016 Headache 001-001 2/4/2016 Vomiting    SUBJID VISITDT TIMEPT CYCLE DAY 001-001 1/4/2016 C1D1 1 1 001-001 1/11/2016 C1D8 1 8 001-001 1/18/2016 C1D15 1 15 001-001 1/25/2016 C2D1 2 1 001-001 2/1/2016 C2D8 2 8 001-001 2/8/2016 C2D15 2 15   From the info above, I can eyeball that the first 2 events occurred in Cycle 1 while the third event occurred in Cycle 2.  Would SQL help me find those proper placements?  Currently, I am taking the visit table and transposing it so that there is a column for each visit (just keeping the CxD1 dates).  Then I merge it in with the AE table (see example table below) and loop through every visit column and check if AEDT is between [C1D1 and C2D1) or [C2D1 and C3D1) (and so on for each cycle visit available).  This seems very ineffective.  Is there a better way?   SUBJID AEDT TERM C1D1 C2D1 001-001 1/7/2016 Nausea 1/4/2016 1/25/2016 001-001 1/9/2016 Headache 1/4/2016 1/25/2016 001-001 2/4/2016 Vomiting 1/4/2016 1/25/2016   ... View more

I'm using GTL to make several bubble plots.  I am making a different plot for a different adverse event where the bubbles are color coded by AE grade.  When I subset my data to certain AE terms, not all grades are represented.  For example, when I run the Fatigue plot, Grade 1, Grade 2, and Grade 3 show up, but Grade 4 is not in the legend because there are no Grade 4 Fatigue in my data.  There are some graphs that only show Grade 1 in the legend.  I just want the Grade 4 to appear in the legend so that all legends are consistent across all plots.   Is there a way in the DISCRETELEGEND statement to specify what values to display (i.e., display all 4 grades)?  I am sure I have done this in the past, but I can't find any sample code.  Google is also not being much of a help.  I am trying to not run some kind of massive shell table that has all possible combinations of subject ID, AE term, and grade level and then merge in my actual data to have a bunch of missing values.  Can I avoid doing that? ... View more

I am not very good with PROC TEMPLATE.  I have a general idea of what I need to do, but I can't seem to get my changes to take hold, so I'm reaching out to the SAS gurus.   I usually use the LISTING style when producing graphs in RTF format through ODS.  When I use a GROUP= option, the style defaults to making the first category blue, the second red, and the third green.  I would like to swap the red and green.   I know there is a GraphColors class, but there are so many options in there that I'm not sure what I'm supposed to be chainging.  I don't know the difference between GDATA, GCDATA, GNDATA, GNCDATA, etc.  Can someone show me a snippet of code that can swap (or even change) colors in a predefined style template?  And maybe as an added bonus, explain the difference between all the GDATA options? ... View more