Seems to work for me.  I am using SAS 9.2 (TS2M3).  Here is the SGPLOT image.  Also works for SGPANEL. ... View more

Use "Courier New" ... View more

Here is an example of what Dan is suggesting: https://sites.google.com/site/smatange/ods-graphics/panel-by-month-and-day ... View more

SAS/GRAPH procedures and annotate will continue to be supported in SAS 9.3 and beyond.  There is no plan to discontinue these features.  While it will be clear to the reader that many new developments in graphics are happening in the ODS Graphics arena, we are also working on new graph procedures in the SAS/GRAPH product for SAS 9.4.  ... View more

The same graph can be done using cluster groups in the SAS 9.3 SGPLOT procedure.  See attached image and code. ... View more

This will get much easier in SAS 9.3, but in the meantime you can use GTL to create any combination of bar and line you need.  If your data is organized in columns like shown in the sas program (attached), then the GTL code will allow you to create the adjascent grouped bars with overlaid lines as shown. ... View more

This option was introduced with SAS 9.2 (TS2M3). With TS2M0, you have to use the SGEDIT ON command in th command line. This has to be done with the Results Window in focus. Now when you run a ods graphics task (SGPLOT or STAT procedure), editable SGE files will be created and listed in the Results window. You can turn SGEDIT OFF the same way. ... View more

No, there is no way to force the curve label below the reference line in a general purpose way. If this is a custom graph, you could create extra X, Y and Label columns with one observation and use a scatter plot as follows. : scatterplot x=x y=x / markercharacter=label markerattrs=(size=0). The label will be centered on (x, y). Adjust X and Y to get the label where you want. Turn off the ReferenceLine label. ... View more

MarkerSize changes from GraphData1-N are not supported. A work around for some cases (where number of groups is known) is to transpose your data into columns, and use a separate scatter statement for each column with MARKERATTRS=(SIZE=15) in the syntax. If it is important that groups support marker size, please call it in to SAS Tech Support. ... View more

You are mostly there. If you really want gray scale colors, you can use style=journal on the ods listing destination for consistent colors. If the "Weeks" is really a axis label, you can leave it as is and get rid of the extra Entry statement ... View more

Just for my information...Did you go with overlaid bars (Proportion2) or side-by-side (Proportion3)? ... View more

Yes, we have a reported defect with mixing of character and numeric data as I mentioned earlier. The work around is to make the X axis data for all plots as discrete (they will become discrete anyway once the defect is fixed). Or, put LAYOUT OVERLAY around each BLOCKPLOT and force the axis to DISCRETE as shown below. You also have to set some DISPLAY options to get the look you want. layout overlay/xaxisopts=(type=discrete display=none) walldisplay=none; ... View more

Note: Bar widths for the 2nd and 3rd bars are reduced so you can see all 3 bars. If you don't set barwidth, the bars will occlude the previous ones. Alternatively, you can also set discrete offset to set the bars side-by-side: barchart x=window y=grppct1 / name='grp1' barwidth=0.2 discreteoffset=-0.2; barchart x=window y=grppct2 / name='grp2' barwidth=0.2; barchart x=window y=grppct3 / name='grp3' barwidth=0.2 discreteoffset=0.2; ... View more

I believe it is the interaction between linear and discrete data between the Bar and the Block plot. StepPlot by default assumes interval data, just like block plots. So thay are all consistent. But when you change the top plot to BarChart, that becomes the primary plot, and the axis is assumed to be discrete. Then, the block plot using the same numeric data cannot draw on the discrete axis. It should, but there may be a bug. If you use a character X varialbe on all plots like the "Window", that works . Or, if you change "Order" to character, that will work too. ... View more