We have an advanced user that suggested to add add Nomogram，DCA decision curve, and NRI in predictive modeling. These functions are now used by R and the users are expecting to use them in JMP as well, especially the medical users.
@mia_stephens
Hi Mia
Thank you for paying attention to my suggestion.
I have asked the SE to program an add-in by combing JMP with R. The JSL is as followed:
// Expressions List
LoadExprs = expr(
InitR = expr(
R Init();
R Submit( "
library(survival)
library(rms)
" );
);
MakeNomoPlot = expr(
yy = yvar << Get Items;
cc = cvar << Get Items;
xx = xvar << Get Items;
R send(yy);
R send(cc);
R send(xx);
dt = Current Data Table();
R Send(dt);
R Submit(
EvalInsert("\[
dd<-datadist(dt)
options(datadist="dd")
t1 <- yy[1]
t2 <- cc[1]
t3 <- paste(xx, collapse = "+")
formula=eval(parse(text=paste("Surv(",t1,",",t2,")~",t3)))
f<-cph(formula,data=lung,x=TRUE,y=TRUE,surv=TRUE)
survival<-Survival(f)
survival1<-function(x)survival(365,x)
survival2<-function(x)survival(730,x)
nom<-nomogram(f,fun=list(survival1,survival2),
fun.at=c(0.05,seq(0.1,0.9,by=0.05),0.95),
funlabel=c('1 year survival','2 year survival'))
plot(nom)
]\"));
);
CapturePlot = expr(
JMP_plot = R Get Graphics(png);
New Window("Imported Plot", Picture Box(JMP_Plot));
);
RunClose = expr(
inwin << Close Window; // close window
R Term( ); //Close r connection
);
);
LoadExprs; //load all the expressions that reside at the end of the file
InitR; //Start R Connection and load libraries
// Prompt for data table if none is open
if(IsEmpty(Current Data Table()), dt = Open(), dt = Current Data Table());
//Draw Window and Create Fields and Buttons
inwin = New Window( //Draw New Window and name inwin
"列线图", //Window Title
Panel Box("选择列", //Draws visible box
H List Box( //Invisible box for layout - glues horizontally
clb1 = Col List Box( all ), //Column listbox, with name clb1 so we can access later
// Draw First Lineup Box (invisible, provides arrangement) for Selecting Columns
Lineup Box( N Col( 2 ), //with 2 columns indicated, each pair will span a row
//First Row
Button Box( "Y,事件时间", yvar << setitems( clb1 << getselected ) ),
yvar = Col List Box( minitems( 1 ), maxitems( 1 ), nlines( 1 ) ),
//Second Row
Button Box( "删失", cvar << setitems( clb1 << getselected ) ),
cvar = Col List Box( minitems( 1 ), maxitems( 1 ), nlines( 1 ) ),
//Third Row
Button Box( "X,因子", xvar << setitems( clb1 << Get Selected ) ),
xvar = Col List Box( minitems( 1 ), maxitems( 100 ), nlines( 10 ) ),
),
// Draw Second Lineup Box for Buttons
Lineup Box( N Col( 1 ),
Button Box("确定", MakeNomoPlot),
Button Box("在JMP窗口中显示", CapturePlot),
Button Box("关闭", RunClose )
) //close Lineup Box
) //close H List Box
) //close Panel box
); //close New WindowThank you for the information. @Clare, is there a script or add-in you can share?
Here is what I found on Wikipedia. I think this link describes its use as well. Is that correct @Clare
Hi Mia
Nomogram，DCA decision curve, and NRI is proposed as advanced and new features in predictive modeling by our senior JMP user in a hospital. He teaches the doctors in hospital how to use JMP. These new functions can be done in R for the moment.
i will share about the nomogram first this time.
Nomogram is commonly used in the medical researching and papers and is very popular among the medical workers.
I did a survey months ago among the doctors in the hospitals. Almost 80-90% of doctors need to use the nomogram in their papers. I asked china SE to program JSL by JMP And R and shared with the customers. But the customers still want to use JMP to make it directly.
If we could add the nomogram to JMP,JMP would become more competitive among those medical workers.
Can you please provide more details? Links to references, links to other requests (if available), an example, screenshots,...
Thanks!
