我的数据
> str(dataCCLM)
'data.frame': 78 obs. of 10 variables:
$ ID : Factor w/ 39 levels "A1","A10","A11",..: 1 1 12 12 23 23 34 34 35 35 ...
$ species : Factor w/ 1 level "c": 1 1 1 1 1 1 1 1 1 1 ...
$ d2H : num -91.5 -72.2 -66.3 -85.6 -57 ...
$ tissue : Factor w/ 2 levels "feather","liver": 2 1 2 1 2 1 2 1 2 1 ...
$ d2HS : num -72.2 -72.2 -85.6 -85.6 -76.8 ...
$ d2HL : num -91.5 -91.5 -66.3 -66.3 -57 -57 -72 -72 -87.8 -87.8 ...
$ guild : int 1 1 1 1 1 1 1 1 1 1 ...
$ numerical.day: int 8 8 9 9 21 21 22 22 23 23 ...
$ Elevation.m. : int 372 372 352 352 115 115 39 39 39 39 ...
$ locality : Factor w/ 13 levels "Bosque Cachil",..: 5 5 6 6 7 7 7 7 7 7 ...
>
我正在尝试创建ID为随机因子的线性混合效果模型
mod.EL.OD <-lmer(d2H ~ Elevation.m. + numerical.day + (1|ID), data = dataCCLM, REML=F)
summary(mod.EL.OD)
我不断收到此缩放错误：
> mod.EL.OD <-lmer(d2H ~ Elevation.m. + numerical.day + (1|ID), data = dataCCLM, REML=F)
Warning message:
Some predictor variables are on very different scales: consider rescaling
> summary(mod.EL.OD)
Linear mixed model fit by maximum likelihood ['lmerMod']
Formula: d2H ~ Elevation.m. + numerical.day + (1 | ID)
Data: dataCCLM
AIC BIC logLik deviance df.resid
649.6 661.4 -319.8 639.6 73
Scaled residuals:
Min 1Q Median 3Q Max
-3.5083 -0.5624 0.0381 0.6384 2.9054
Random effects:
Groups Name Variance Std.Dev.
ID (Intercept) 8.72 2.953
Residual 204.70 14.307
Number of obs: 78, groups: ID, 39
Fixed effects:
Estimate Std. Error t value
(Intercept) -77.950113 3.770435 -20.674
Elevation.m. -0.001115 0.001161 -0.961
numerical.day 0.056195 0.018016 3.119
Correlation of Fixed Effects:
(Intr) Elvt..
Elevatin.m. -0.286
numericl.dy -0.731 -0.283
fit warnings:
Some predictor variables are on very different scales: consider rescaling
>
我已经关注了其他一些论坛建议并使用此代码来扩展我的数据：
pvars <- c( "d2H",# Scaling numeric parameters
"Elevation.m.","numerical.day",
"d2HL","d2HS")
datscCC <- dataCCLM
datscCC[pvars] <- lapply(datscCC[pvars],scale)
然而我仍然得到了警告...... 这个模型不适合我的数据，还是我可以采取其他重新缩放步骤？