Okay, the output for the table you have included above is usually for a forest plot. Maybe that is the plot you are after? An example of a forest plot can be found here: http://blogs.sas.com/content/graphicallyspeaking/2013/01/06/forest-plot-with-sas-9-3/ In the figure you want to produce, you have time on the x-axis. Have you also outputted a dataset that includes the time and the hazard ratio's? ... View more

Hello, I'm guessing you can use the Series statement within SGPLOT to get the plot that you want. Can you please include the &infile dataset or the HR&infile dataset so that I can verify? ... View more

If  you have outputted the corresponding dataset then you could use SGPLOT with three SERIES statements to obtain the Hazard Ratio vs Covariate and the 95% confidence limits. You could also plot the reference line using the REFLINE statment. Have you also checked to see if you can get a plot similar to the one you want directly from using the PROC PHREG procedure? ... View more

Hi Lisa, It's difficult for me to visualise your graph. Can you show an example of it please? Also please let us know the SAS Version you have and include an example of the dataset. It may also be better to post this question in the SAS/GRAPH and ODS Graphics community, if you haven't already. Thank you. Kriss ... View more

Hi Andreas, A quick overview of the differences between Discreteattrmap and the GraphData, is that when you use the Discreteattrmap and you specify that "Males" should have Blue markers, and "Females" should have Pink markers, then that would always be the case, regardless of drop outs of the data or the sorting order of the data. Now, the attributes that are selected when you use GraphData are based on the order of the data, and so Females, would obtain the first group GraphData1, which by default is a Blue colour, and Males would obtain the values specified in the second group which by default is a Red colour. However if there where not any Females in the dataset, then Males would now obtain the values specified in GraphData1 which is the Blue colour. With attributes maps, you need to use the Group = statement to assign the map. The Blockplot has not got the group statement so I guess that is why you can not use attribute maps with that plot. Thanks. ... View more

Hello, try adding the following statement before SGPANEL, but instead of just using "C:", use a location where you have access to write to.   ods listing gpath = "C:"; Thanks. ... View more

Hello, You could use the discrete offset options or the blockplot statment which are explained on the Graphically speaking blog linked below. http://blogs.sas.com/content/graphicallyspeaking/2011/11/12/graphs-with-axis-aligned-statistics-tables/ http://blogs.sas.com/content/graphicallyspeaking/2011/11/10/it-pays-to-be-discrete/ Although, with these methods and because you want to produce a boxplot you will need to use GTL instead of SGPLOT. You will also need to create a column that contains the median of each year. Please let me know if you have any questions. Thanks. ... View more

Great post! Thanks for this! ... View more