[R] Code OK - Predicting X from Y in four-parameter fit (SSfpl) -
Klaus Jensen
KSA at ssi.dk
Thu Jul 8 16:20:59 CEST 2004
[Sorry for any inconvenience - Code is OK now]
While analyzing titration-data using four parameter fit, I need to
predict single values from the model: I need to predict both ways (ie:
X->Y & Y->X):
Example:
library(nls)
cramp<-c(33,100,300,900,2700,8100,24300,72900)
myo<-c(2.7130,2.6790,1.5255,0.7675,0.3670,0.2150,0.1575,0.1400)
titration<-data.frame(cramp,myo)
fi<-nls(myo~SSfpl(log(cramp),A,B,xmid,scal),data=titration)
SSfpl(log(2000), coef(fi)[1],coef(fi)[2],coef(fi)[3],coef(fi)[4])
-> 0.3689212
This is an example of predicting single values X->Y
Here is my question:
How do I predict the other way around Y->X ??
Any help is appreciated
Best regards,
Klaus
State Serum Institute, Copenhagen
More information about the R-help
mailing list