Displaying 1 result from an estimated 1 matches for "femaledata".
2013 Jan 10
2
piece-wise linear regression nls function
...pt. I am using the nls code below, and get the following error message:
Error in nls(FM ~ blow * BMIJS + bhi * sapply(BMIJS - knot, max, 0), start = list(knot = 25, :
singular gradient
nls code:
test <- nls(FM~blow*BMIJS+bhi*sapply(BMIJS-knot,max,0),start=list(knot=25,blow=1,bhi=1),data=FeMaleData)
summary(test)
greatly shortened version of my data (the full data set has 450 records)
FM BMIJS
2 55.878 40.57273
4 34.270 27.76939
5 20.123 21.73818
6 19.320 19.71203
9 49.701 43.55356
10 51.188 37.84742
11 46.753 37.71003
13 65.079 37.23438
14 37.097 36.81806
15 30.625 29....