calibrate curve for binary logistic model in rms packages

Question

[wonderfulzhu1985]

I do an binary logistic regression to construct an nomogram. And then do an calibrate curve to Resampling Model Calibration. the code is list below:

library(rms)
f1 <- lrm(Recurrence_1_year~AFP_Categ_200+Tumor_Diameter_Categ+Macro_Tumor_Thrombus,
               data = Training_nom,
               x=TRUE,
               y=TRUE)     
plot(cal,
      subtitles=FALSE,                    
      conf.int=TRUE,                              
      cex.subtitles=.75,                  
      riskdist=FALSE,                     
      add=FALSE,
      scat1d.opts=list(nhistSpike=1000), 
      par.corrected=list(col="red",lty=4,lwd=4))

when i run the code, many warnings present in the interface of R-studio, those warnings are like those:

1: In plot.window(...) : "conf.int" not graphic paremeters
2: In plot.window(...) : "cex.subtitles"not graphic paremeters
3: In plot.window(...) : "riskdist"not graphic paremeters
4: In plot.window(...) : "add"not graphic paremeters
5: In plot.window(...) : "par.corrected"not graphic paremeters
6: In plot.xy(xy, type, ...) : "conf.int"not graphic paremeters
7: In plot.xy(xy, type, ...) : "cex.subtitles"not graphic paremeters
8: In plot.xy(xy, type, ...) : "riskdist"not graphic paremeters
9: In plot.xy(xy, type, ...) : "add"not graphic paremeters
10: In plot.xy(xy, type, ...) : "par.corrected"not graphic paremeters
11: In axis(side = side, at = at, labels = labels, ...) : "conf.int"not graphic paremeters
12: In axis(side = side, at = at, labels = labels, ...) : "cex.subtitles"not graphic paremeters
13: In axis(side = side, at = at, labels = labels, ...) : "riskdist"not graphic paremeters
14: In axis(side = side, at = at, labels = labels, ...) : "add"not graphic paremeters
15: In axis(side = side, at = at, labels = labels, ...) :  "par.corrected"not graphic paremeters
16: In axis(side = side, at = at, labels = labels, ...) : "conf.int"not graphic paremeters
17: In axis(side = side, at = at, labels = labels, ...) : "cex.subtitles"not graphic paremeters
18: In axis(side = side, at = at, labels = labels, ...) : "riskdist"not graphic paremeters
19: In axis(side = side, at = at, labels = labels, ...) : "add"not graphic paremeters
20: In axis(side = side, at = at, labels = labels, ...) : "par.corrected"not graphic paremeters
21: In box(...) : "conf.int"not graphic paremeters
22: In box(...) : "cex.subtitles"not graphic paremeters
23: In box(...) : "riskdist"not graphic paremeters
24: In box(...) : "add"not graphic paremeters
25: In box(...) : "par.corrected"not graphic paremeters
26: In title(...) : "conf.int"not graphic paremeters
27: In title(...) : "cex.subtitles"not graphic paremeters
28: In title(...) : "riskdist"not graphic paremeters
29: In title(...) : "add"not graphic paremeters
30: In title(...) : "par.corrected"not graphic paremeters

All the parameters list in the warnings was also list in the introduction of the rms packages, which also played an important role in getting a adjustment for the plot. So I wonder to know there is any mistake when i do the calculate. And some one encounter the same problem. Could you share your method to solve the problem. Thanks very much.

Answer 1

[Kevin Blighe]

Hey,

You forgot the calibrate() function. Try this, instead:

library(rms)

f1 <- lrm(Recurrence_1_year~AFP_Categ_200+Tumor_Diameter_Categ+Macro_Tumor_Thrombus,
               data = Training_nom,
               x=TRUE,
               y=TRUE)     

cal <- calibrate(f1, group=Recurrence_1_year)

plot(cal,
      subtitles=FALSE,                    
      conf.int=TRUE,                              
      cex.subtitles=.75,                  
      riskdist=FALSE,                     
      add=FALSE,
      scat1d.opts=list(nhistSpike=1000), 
      par.corrected=list(col="red",lty=4,lwd=4))

Kevin