将幂律函数拟合到 R 中的数据

Question

我正在尝试使用此处的逐步方法将幂律函数拟合到我的数据

https://www.statology.org/power-regression-in-r/

但我不清楚错误

```Error in lm.fit(x, y, offset = offset, singular.ok = singular.ok, ...) : NA/NaN/Inf in 'x'```

我在这里查了错误https://community.rstudio.com/t/error-error-in-lm-fit-x-y-offset-offset-singular-ok-singular-ok-na-nan-inf-in-y/30481

它说我需要确保它是我所做的 as.numeric，但我仍然得到错误。我怎样才能对此拟合幂律并得到方程？

 dput(data)
 structure(list(Depth = c(0, 1.12941645231161, 2.25883290462321, 3.38824935693482, 
 4.51766580924643, 5.64708226155804, 6.77649871386964, 7.90591516618125, 
 9.03533161849286, 10.1647480708045, 11.2941645231161, 12.4235809754277, 
 13.5529974277393, 14.6824138800509, 15.8118303323625, 16.9412467846741, 
 18.0706632369857, 19.2000796892973, 20.3294961416089, 21.4589125939205
 ), Biomass = c(2.38380058571429, 2.38351444250963, 2.38322874604698, 
 2.38294349557974, 2.38265869045564, 2.38237435716909, 2.382090559166, 
 2.38180732244924, 2.38152464168852, 2.38124252252291, 2.38096100921642, 
 2.38068013931138, 2.38039991126563, 2.38012032354094, 2.3798413897966, 
 2.37956313811693, 2.37928557316563, 2.37900869200497, 2.37873250170607, 
 2.37845703591721)), row.names = c(NA, 20L), class = "data.frame")


 data$Depth <- as.numeric(data$Depth)
 data$Biomass <- as.numeric(data$Biomass)

 model <- lm(log(data$Biomass) ~ log(data$Depth))
 summary(model)

Answer 1

您的问题出在日志上。

summary(data)
     Depth           Biomass     
 Min.   : 0.000   Min.   :2.378  
 1st Qu.: 5.365   1st Qu.:2.380  
 Median :10.729   Median :2.381  
 Mean   :10.729   Mean   :2.381  
 3rd Qu.:16.094   3rd Qu.:2.382  
 Max.   :21.459   Max.   :2.384

由于 Depth 的值为零，因此 log(data$Depth) 的值等于 -Inf

将幂律函数拟合到 R 中的数据

Fitting power law function to data in R

r

equation

lm