hello,
can anyone help with this:
###########################################################
###data: measurments (response = trans) run several times at the same
predictor value level (press)
por<-data.frame(list(structure(list(run = structure(c(1L, 1L, 1L, 1L, 2L,
2L,
2L, 2L, 3L, 3L, 3L, 3L), .Label = c("1", "3",
"4"), class = "factor"),
press = c(15, 21, 24, 29.5, 15, 21, 24, 29.5, 15, 21, 24,
29.5), tr_rel = c(1, 0.459454191, 0.234697856, 0.135282651,
1, 0.853283066, 0.306314797, 0.186302231, 1, 0.42980063,
0.103882476, 0.086463799), tr = c(513, 235.7, 120.4, 69.4,
318.3, 271.6, 97.5, 59.3, 476.5, 204.8, 49.5, 41.2)), .Names =
c("run",
"press", "tr_rel", "tr"), row.names = c(NA, -12L),
class = "data.frame")))
attach(por)
###nls:
summary(mod1<-nls(tr ~ SSlogis( log(press), Asym, xmid, scal)))
press_x <- seq(15, 30, length = 100)
####predict() somehow doesn't recognize the list and returns values
####as if list argument was absent:
predict(mod1,list(Pressure=press_x))
####consequently plotting of fitted line doesn't work either:
plot(press,tr)
lines(press_x,predict(mod1,list(pressure=press_x)))
###########################################################
thanks a lot in advance for any hints -
greetings,
kay
--
View this message in context:
http://n4.nabble.com/nls-predict-problem-tp1690561p1690561.html
Sent from the R help mailing list archive at Nabble.com.
hello,
sorry to have hassled you with this - in the end it is all quite simple and
after digging deeper into the help files, also of predict.lm, i finally
managed to do it.
here is the worked example with solution for anyone beeing clumsy as i was:
###########################################################
por<-data.frame(list(structure(list(run = structure(c(1L, 1L, 1L, 1L, 2L,
2L,
2L, 2L, 3L, 3L, 3L, 3L), .Label = c("1", "3",
"4"), class = "factor"),
press = c(15, 21, 24, 29.5, 15, 21, 24, 29.5, 15, 21, 24,
29.5), tr_rel = c(1, 0.459454191, 0.234697856, 0.135282651,
1, 0.853283066, 0.306314797, 0.186302231, 1, 0.42980063,
0.103882476, 0.086463799), tr = c(513, 235.7, 120.4, 69.4,
318.3, 271.6, 97.5, 59.3, 476.5, 204.8, 49.5, 41.2)), .Names =
c("run",
"press", "tr_rel", "tr"), row.names = c(NA, -12L),
class = "data.frame")))
attach(por)
summary(mod1<-nls(tr ~ SSlogis( log(press), Asym, xmid, scal)))
predict(mod1, x_press)
x_press<-data.frame(press=seq(from=5,to=35,by=0.5))
y_pred<-data.frame(tr=predict(mod1, x_press))
plot(press,tr, xlim=c(10,30))
lines(cbind(x_press,y_pred))
###########################################################
greetings,
kay
--
View this message in context:
http://n4.nabble.com/nls-predict-problem-tp1690561p1692061.html
Sent from the R help mailing list archive at Nabble.com.