I'm able to generate a nice plot of Product Limit Survival Estimate via PROC LIFETEST, and even capture it in an .rtf file.  I'm wondering if there are any ways to customize it:   1) add titles and footnote  2) remove tabular output that LIFETEST generates from the .rtf file that contains the plot.     Maybe the answer is to write the plotting code myself instead of relying on generic LIFETEST output.  I tried extracting the time and survival rate data using this statement:     ods output survivalplot=zz; * capture the time and survival rates in a data set **;   But I tried a simple GSPLOT line plot and it didn't have that squared-off look that a survival plot usually has.   It seems like this is a common type of display, so it's frustrating that I can't figure it out.  Anybody have any ideas?   Thanks,   ...dave ... View more

  I am trying to calculate a survival rate and corresponding 95% CI from PROC LIFETEST.  Is there an ODS data set I can grab that has these?   As I understand it, the survival rate should be the rate at the last observation, right?  If so, my survival rate is .0938.  But how can I get an 95% CI to go along with that?   Thanks   0.000 1.000000012Product-Limit Survival EstimatesAVAL   Survival Failure Survival Standard Error NumberFailed NumberLeft 50.000   0.9167 0.0833 0.0798 1 11 51.000   0.8333 0.1667 0.1076 2 10 52.000   0.7500 0.2500 0.1250 3 9 53.000 *       3 8 106.000   0.6563 0.3438 0.1402 4 7 114.000   0.5625 0.4375 0.1482 5 6 120.000   0.4688 0.5313 0.1503 6 5 126.000   0.3750 0.6250 0.1466 7 4 162.000   0.2813 0.7188 0.1367 8 3 163.000   0.1875 0.8125 0.1190 9 2 164.000   0.0938 0.9063 0.0891 10 1 168.000 *       10 0   ... View more

Is there any way to suppress all the on-screen "help" in SAS Enterprise?  It seems that with every key stroke, boxes of text pop up and disappear.  Sometimes It's even inserting key words and and altering what I type, and the boxes are always in my way.  Today all the flashing is actually making me feel nauseous, although it might be the bad coffee I'm drinking.   Thanks! ... View more

I'm trying to specify values for the y axis from 0 at the top to -2 where it meets the x axis.  I can't seem to please Mr. SAS.   When I do this: axis1 label=(a=90 h=12pt 'Threshold (log10 sec)') order=(0 to -2 by 0.2) /*(-2 to 0.0 by 0.2)*/; axis2 label=(h=12pt 'Session') order=(1 to 10 by 1) minor=none;  I get this in the log: 23 GOPTIONS ACCESSIBLE; 24 axis1 label=(a=90 h=12pt 'Threshold (log10 sec)') 25 order=(0 to -2 by 0.2) /*(-2 to 0.0 by 0.2)*/; WARNING: Order= clause on the axis statement is not valid. The default axis values will be used. 26 axis2 label=(h=12pt 'Session') order=(1 to 10 by 1) minor=none; When I do this, the log is clean but instead of printing '0' at the top, it prints "-2..7756E-16" in place of the "0" (all the other numbers on the scale appear as I want them).    axis1 label=(a=90 h=12pt 'Threshold (log10 sec)') order=/*(0 to -2 by 0.2)*/ (-2 to 0.0 by 0.2); axis2 label=(h=12pt 'Session') order=(1 to 10 by 1) minor=none;   So what's the secret?  I simply want this along my Y axis:   0   -.2   -.4   -.6   -.8   -1.0   (all the way to -2.0)   ... View more

Hi, I'm trying to generate .cgm outputs for all my plots.  I'm finding that When I use PROC GPLOT or PROC BOXPLOT, I'm able to generate .cgm files.   But when I use SGPLOT or SGPANEL, no .cgm file appears.  There is no error message, but also no file.  I can see the plot in the results tab of Enterprise.   It can't be the case that SGPLOT and SGPANEL can't do a .cgm file, can it?  My processing statements are the same for both.     filename gout 'myfile.cgm';   goptions device = [cgmofml cgm cgmof97l] gsfname = gout gsfmode = replace;  * tried all three devices, none work.   proc gspanel/gsplot] . . .   Thanks for any insight.  You all have been so helpful in the past week to me and it's much appreciated! ... View more

As you know by now, I'm just learning about plotting!   I have developed a line plot of a mean across timepoints, with bars for 95% confidence interval. (PROC GPLOT) I'm trying to create a very simple legend, simply indicating that the red line is treatment group 1 and the blue line is treatment group 2.  Everything I try generates stuff that I don't want.    Is there an easy way to populate the legend with something simple like this?   --- Placebo      ---- Study Drug     Here is my code (I found it online and it does exactly what I need).  Mean1 is the mean per session, Thresh1 contains LCM and UCM.   proc gplot data=prep5; plot thresh1*session mean1*session thresh2*session mean2*session / overlay haxis=axis1 vaxis=axis2 /*legend=legend1*/; title1 h=8pt color=black font='courier' J=L "&titlel1"; title2 h=8pt color=black font='courier' J=L "Protocol: &titlel2 Page 1 of 1"; title5 h=8pt color=black font='courier' J=C "&titlel5"; title6 h=8pt color=black font='courier' J=C "&titlel6"; title7 h=8pt color=black font='courier' J=C "&titlel7"; run; ... View more