I work for an organization that conducts clinical trial research, and our organization has a standardized powerpoint template for the reporting of all of our results.  I've been tasked with coming up with programming to take advantage of the ODS PowerPoint destination to streamline generation of slides with tables of statistics.  I'm having trouble figuring out how to change the color of fonts to match the odd colors that our graphics department has chosen for the slide template.  I can't find any sample programming showing exactly how to change the font colors.  The closest I've gotten is Tim Hunter's 2013 SAS Global Forum paper (https://support.sas.com/resources/papers/proceedings13/041-2013.pdf) that mentions but doesn't demonstrate how to make this particular change.   Below is what I think should work.  The colors all match up with the colors in my organizations template.  But what I get is a slide that is entirely black.  The background is black, all the text is black...everything is black.  Can someone show me where I've gone wrong?   I am using 64 bit Windows 7, SAS 9.4, TS Level 1M2, and have Office 2010. Thanks!   ods path work.templat(update) sasuser.templat(update) sashelp.tmplmst(read); ods path (prepend) work.templat(update); proc template; define style ppt; parent=styles.powerpointlight; scheme 'ODS Light' / dark1 = #cx565656 light1 = #cxffffff dark2 = #cx565656 light2 = #cxffffff accent1 = #cx98012E accent2 = #cx565656 accent3 = #cxffffff accent4 = #cxffffff accent5 = #cxffffff accent6 = #cxffffff heading_font = ("Arial",32pt) body_font = ("Arial",20pt); end; run; ods powerpoint style=ppt; title "Children 14+"; proc report data = sashelp.class nowd; columns Name sex age height weight; where age ge 14; run; ods powerpoint close; ... View more

I'm in the process of overhauling standard survival graph programming where I work, getting away from PROC GPLOT and massive ANNOTATE datasets to add additional information which the statisticians want to have included.  I am moving instead to PROC TEMPLATE and SGRENDER-based programming instead since it is so much more simple.  I'm having some issues, though, with the PNG files that SGRENDER is producing when I import them into powerpoint.  Our group has a standard slide template, and it is expected that the survival graph will be transparent, allowing the slide template to show through.  However, the PNG file is not transparent - it's a big white box.  I've tried specifying OPAQUE=FALSE in the statgraph template I've made, as well as augmenting the DEFAULT style template to use in the ODS LISTING line I have right before the PROC SGRENDER statement.  None of this is making the PNG file be transparent when imported into powerpoint.  I'm not terribly adept at working with style templates, so I could easily have messed that up.  The following are my graph & style templates, followed by the SGRENDER statement. Any tips on how to get that PNG file transparent in the powerpoint file and have it transparent so the powerpoint slide template shows through?     proc template;         define statgraph SURVIVAL_GRAPH;             begingraph;                 layout lattice / rows=1 columns=1 opaque=false;                     layout overlay / xaxisopts=(linearopts=(tickvaluelist=(0 3 6 9 12 15 18 21 24) viewmax=24))                                      yaxisopts=(linearopts=(tickvaluelist=(0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0) viewmin=0));                         stepplot x=time y=survival /  group = stratum name='survgraph';                         scatterplot x=time y=censored /  group=stratum markerattrs=(symbol=plus);                         innermargin / align=bottom;                             blockplot x=tatrisk block=atrisk / class="stratum" display=(values label) valueattrs=(size=8)                                       valuehalign=start;                         endinnermargin;                         discretelegend 'survgraph' / location = inside halign=center valign=bottom;                     endlayout;                     columnheaders;                         layout overlay / pad=(top=5);                         layout gridded / columns=7 border=true;                             entry " ";                             entry halign = center "# of Patients";                             entry halign = center "Dead";                             entry halign = center "Censored";                             entry halign = center " ";                             entry halign = center "Median Survival (95% CI)";                             entry halign = center "Hazard Ratio (95% CI)"; *...ADDITIONAL LINES NOT SHOWN....;                         endlayout;                         endlayout;                     endcolumnheaders;                 endlayout;             endgraph;         end;     run; proc template;     define style journal_transparent;       parent = Styles.Journal;         style graphbackground from graphbackground/ transparency = 1.00;         style GraphLegendBackground from GraphLegendBackground/ transparency = 1.00;         style GraphCharts from GraphCharts /    transparency = 1.00;         style GraphLegendBackground from GraphLegendBackground / transparency = 1.00;         style GraphDataDefault from GraphDataDefault / transparency = 1.00;         style graphdata1 from graphdata1 / transparency = 1.00;         style graphdata2 from graphdata2 / transparency = 1.00;     end; run; ods graphics / reset imagename = "survival graph" IMAGEFMT=png ANTIALIAS=off; option orientation = landscape; ods listing image_dpi = 300 GPATH="&directory\sas output" style=journal_transparent;      proc sgrender data=_survivalplotdata template=SURVIVAL_GRAPH;      run; ... View more

I'm trying to use the ODS Output dataset ParameterEstimates from the PHREG procedure, and I'm having an issue where it appears that the variable "Parameter" only has a length of 20, so it's truncating any parameter entered into the model with length > 20.  This is problematic for me since, amongst other variables, I'm trying to use three interaction term variables named volume_gtv_rt_tech_d1, volume_gtv_rt_tech_d2, and volume_gtv_rt_tech_d3.  The ParameterEstimates dataset has the value volume_gtv_rt_tech_d in the Parameter variable for all 3 terms.  My ultimate goal is to combine this data with some PROC FREQ data so I can make a table for my statistician that includes the # of failures within each covariate grouping, and this is now rendered nearly impossible with truncating data. Has anyone else come across this issue?  Other than renaming my variables (which is problematic since I do not have complete control over them), is there some other work-around?  A random option in PHREG that I'm not aware of?  I'm running 9.2, so maybe there's a fix in the latest release, and I could use this as a bargaining chip to get upgraded? Thanks!! ... View more