我有来自威布尔累积密度分布的累积概率，并希望找到基础函数参数。
数据：
At <- data.frame(cumu = c(0.004722245,0.004740720,0.008768300,0.010900330,0.194195851,0.215752640,0.226863587,0.237534826,0.240753194,0.324707908,0.354541558,0.421462196,0.421879734,0.423690298,0.454939143,0.597567194,0.647963670,0.688549850,0.729575737,0.742892615,0.758127212,0.760495725,0.762376494,0.763717789,0.764586120,0.985689160,0.993160505,0.995931775,1.000000000),yr = c(1,2,4,5,6,7,8,9,10,11,12,13,14,15,16,17,22,32,33,34,35,36,37,38,40,41,42,43,46))
我已经尝试通过nls函数通过MLE进行参数拟合，但是它会抛出错误
test <- nls(At$cumu ~ c * exp(-(At$yr/b)^a), start = list(a = .01, b = .2, c = 1))
错误：
numericDeriv（form [[3L]]，names（ind），env）出错： 评估模型时产生的缺失值或无穷大
我认为问题是错误的起始值。所以我的问题是哪些起始值适合我的数据？
谢谢，
西蒙