Question

我正在尝试在gnls中运行nlme模型（包R），并正在排查错误消息。到目前为止，我的模型是：

Log13.5_Sum.T_gnls <- gnls(Weight ~ Asym/(1+exp((xmid-Age)*(v0 + v1*Sum.T))),
                           data = WTS_w13.5,
                           start = c(coef(Log13.5_gnls), v0= 0.16, v1= -0.0059),
                           na.action=na.omit,
                           control= gnlsControl(nlsTol=0.01))

发出错误：step halving factor reduced below minimum in NLS step

我已经尝试将nlsTol的值按照数量级增量增加到1000，但这会产生同样的问题。

我之前使用SSlogis中的nls解决了类似的问题，以生成起始参数值。这些值告知模型Log13.5_gnls，其系数是此处的起始值。 v0和v1的估算值来自于对问题的记录形式使用线性回归的解决方案，因此我期望它们是合理的。

Data.frame：

structure(list(Nest = structure(c(1L, 17L, 11L, 5L, 6L, 8L, 13L, 5L, 7L, 21L, 9L, 15L, 17L, 4L, 10L, 16L, 20L, 7L, 8L, 18L, 19L, 13L, 14L, 11L, 21L, 6L, 9L, 12L, 17L, 7L, 13L, 2L, 3L, 5L, 7L, 10L, 19L, 22L, 17L, 4L, 21L), .Label = c("WTSN00", "WTSN01", "WTSN02", "WTSN04", "WTSN05", "WTSN06", "WTSN07", "WTSN08", "WTSN09", "WTSN10", "WTSN12", "WTSN13", "WTSN14", "WTSN16", "WTSN18", "WTSN20", "WTSN21", "WTSN23", "WTSN24", "WTSN25", "WTSN26", "WTSN28", "WTSN29"), class = "factor"),
Hatch = structure(c(10957, 16177, 16181, 16165, 16185, 16188, 16181, 16165, 16189, 16189, 16193, 16195, 16177, 16177, 16181, 16200, 16173, 16189, 16188, 16182, 16176, 16181, 16180, 16181, 16189, 16185, 16193, 16177, 16177, 16189, 16181, 16177, 16177, 16165, 16189, 16181, 16176, 16181, 16177, 16177, 16189), class = "Date"),
Age = c(0, 6, 7.5, 8, 8, 8, 8, 8.5, 8.5, 8.5, 9, 9, 9, 9.5, 9.5, 9.5, 9.5, 10, 10, 10, 10, 10.5, 10.5, 11, 11, 11.5, 11.5, 11.5, 11.5, 12, 12, 12.5, 12.5, 12.5, 12.5, 12.5, 12.5, 12.5, 13, 13.5, 13.5),
Weight = c(1.022, 8.1, 8.5, 8.8, 8.8, 9.6, 8.6, 9.7, 11, 9.9, 11.1, 9.9, 12, 10.5, 10.5, 7, 11.2, 11.9, 11.4, 11, 11.9, 11.2, 11.7, 9.1, 12.3, 12.3, 13, 11.6, 13.4, 12.2, 11.1, 12.7, 11.3, 12.2, 12.4, 11.8, 12.9, 11.2, 13.2, 11, 14.1),
Sum.T = c(NA, 313.851904139019, 375.724003527149, NA, 406.89428107332, 463.691957819221, 445.28632226469, NA, 415.486307859544, 429.22166483835, 448.628929837252, 436.477157418387, 480.998397305757, 505.76849873214, 465.046041424436, 529.914690217322, NA, 488.827490583208, 582.676479904347, 570.255184933855, NA, 594.21509635292, 703.48545276177,  559.086005290723, 569.478435100335, 593.644945574203, 581.137438612915, 572.523390933983, 630.703808278039, 584.949067548094, 692.972360805858, 690.676762264357, 698.225577774177, NA, 608.229461789316, 637.728055232582, NA, NA, 722.027054861407, 759.276720390869, 701.235205362319)),
.Names = c("Nest", "Hatch", "Age", "Weight", "Sum.T"),
row.names = c(NA, 41L), class = "data.frame")

有关如何解决此问题的任何想法？

在NLS步骤中理解错误消息＆＃34;步骤减半因子减少到最小值＆＃34;

0 个答案: