.-
help for ^stcoxkm^
.-

Plot of Predicted vs. Observed Survival Curves:                 [STB-35: gr23]
-----------------------------------------------
(J.M.Garrett 12/16/96)

   ^stcoxkm^ [^if^ exp] [ ^,^ ^by(^xvar^) nosh^ow graph_options]


^stcoxkm^ plots the predicted (Cox Proportional Hazards model) vs. observed
    (Kaplan-Meier) survival curves by categories of a nominal x variable. This
    plot can be useful for examining the proportional hazards assumption --
    the closer the observed values are to the predicted, the less likely the
    assumption has been violated. You must have ^stset^ your data before using
    this command; see help @stset@.


Options required
-----------------
  ^by(^xvar^)^ -- nominal or ordinal x variable


Options allowed
---------------    

  ^noshow^ -- prevents ^stcoxkm^ from showing the key ^st^ variables
  graph_options -- most of the options available in ^graph^
   

Examples 
--------

 . ^stset weeks relapse^
 . ^stcoxkm, by(trtment) c(..ll) s(oopp) xlabel^

     Plots the Cox predicted survival curves vs. the Kaplan-Meier observed
     survival curves for weeks to relapse for leukemia patients by separate
     categories of treatment; connects the predicted curves with lines; uses
     "o" for observed values and "+" for predicted values

 . ^stcoxkm, by(trtment) c(JJll) s(iipp) xlabel ylabel(0,.5,1)^

     Repeats the previous example, using a step function and no symbols for
     the observed survival curves

 . ^stcoxkm, by(wbc3cat) c(...lll) s(oooppp)^ 

     Plots predicted and observed survival curves for three cateogories of
     white blood cell count


Also see
--------

    STB:   gr23 (STB-35)
 Manual:   ^[R] st sts graph^
           ^[R] st stcox^
 On-line:  help for @st@, @sts@