I have the same question. ... View more

I managed to upgrade the result a little by using contents instead of page and adding some text in between. However, by using outputfmt=svg, the image size should adapt to the screen used (large screen, laptop screen,...) but this seems not to be the case. ods noproctitle; * with ods graphics / imagefmt=svg the contents will resize to the size of the browser container ; ods html5 style=opal path= "/data/temp" (url=none) body = 'test_body.htm' contents = 'test_contents.htm' /*creates toc*/ frame = 'test_web.htm' /*page = 'test_pagecontent.htm' newfile=page*/ options(svg_mode="inline"); /* setup the graphs */ ods graphics on / reset=all scale=off imagefmt=SVG outputfmt=svg imagemap=on; /* needed for data tips */ ods html5 text= '<h1>Scatter Plots</h1>' ; ods layout gridded rows=1 columns=2 column_gutter=20px; ods region row=1 column=1; ods ods proclabel "weight vs height for male students"; title j=c "weight vs height for male students"; proc sgplot data = sashelp.class description =' ' noopaque; scatter x=weight y=height; where sex='M'; run; ods region row=1 column=2; ods proclabel "weight vs height for female students"; title j=c "weight vs height for female students"; proc sgplot data = sashelp.class description =' ' noopaque; scatter x=weight y=height; where sex='F'; run; ods layout end; ods html5 text = '<br>' ; ods html5 text = '<hr>' ; ods html5 text = '<br>' ; ******************************************; ods html5 text= '<h1>Boxplots</h1>' ; ods layout gridded rows=1 columns=2 column_gutter=20px; ods region row=1 column=1; ods proclabel "boxplots weight"; title j=c "boxplots weight"; proc sgplot data = sashelp.class description =' ' noopaque; vbox weight / category = sex ; run; ods region row=1 column=2; ods proclabel "boxplots height"; title j=c "boxplots height"; proc sgplot data = sashelp.class description =' ' noopaque; vbox height / category = sex ; run; ods layout end; ods html5 close; ... View more

ah yes, that makes sense! I had it before and then removed it when adding groupmc and groupms. ... View more

For the record, the DATALABEL option on bar charts cannot put a label on top of a bar when there are limits enabled. If you try it, we will automatically switch to axis tables internally to display the information (we do this as a convenience). If you know that you want to customize these labels in this situation, it is best to go ahead and use axis tables directly.   Hope this helps! Dan ... View more

The information is very scattered around. For example there is ods rtf that you can use with proc odstext, but then there is also ods tagsets.rtf with ods tagsets.rtf text = to add text. Then there is ods layout for rtf and ods layout for pdf.    I am using mainly following workflow: a proc freq, descriptive statistics with proc tabulate, then graphs, then some statistical model where I want to add the diagnostics panel and tables with lsmeans and contrast statements results and then some text with the used methodology and interpretation of the results. The latest that I tried was ods tagsets.rtf. I played around with the rtf mark-up language and switched then to styles in the form of "^S={ }. It;s difficult to find for example how to change margins, I first changed it in the template, then I found that there is also an ODS option for it. Changing vertical spaces between graphs and texts. Getting titles and captions beneath graphs does also not seem easy, and to obtain a table of contents, you must first open the rtf and click to get actually the title. To find out all these, I had to read already a lot of SAS papers and I still did not find the solutions that I need. Generally I find my way in the user guide, the SAS papers and the blogs, except for this. ... View more

Look through this blog: https://blogs.sas.com/content/graphicallyspeaking/   It has examples on pretty much anything you would ever want to do with graphs/figures.  May take a bit of searching, but it is a very comprehensive source of material. ... View more

  If WEEK is a classification variable, and if you are thinking of REP as the replicating experimental unit for TREATMENT and of PLANTID as subsamples within each REP, and if you want to avoid pseudoreplication, then I concur with the code from Schabenberger and Pierce.   Although you don't derive any denominator df benefits from all those individual plots within a REP, the subsamples still have value: the mean computed over 150 plants is a better metric, with a smaller variance, than the value from a single plant.   ... View more

lvm has posted some really good things regarding adding a constant to values, and then using lognormal and gamma distributions, and what might go wrong.  I know I have been doing that for quite a lot of things.  However, PROC GENMOD does provide zero-inflated techniques for Poisson and negative binomial distributions--I just don't know if it will accept GEE (REPEATED) syntax. Steve Denham ... View more