广义混合效应模型中某个因子的2个水平会丢失吗?

时间:2020-01-27 22:04:49

标签: r categorical-data effects levels mixed

我运行了这个广义混合效应模型

glm1bd <- glmer(Abundance ~ Reason + Region + months.since.spill + bareground.cover +
                  litter.cover + Distance.to.nearest.bush + buildings + roads + (1|Site/Session/Point), warrispill,
                control = glmerControl(optimizer= "bobyqa"),
                family=poisson, nAGQ=1)

summary(glm1bd)

这是我获得的固定效果的输出。

                          Estimate Std. Error z value Pr(>|z|)    
(Intercept)               1.434451   0.154745   9.270  < 2e-16 ***
ReasonIllegal Refinery   -0.440155   0.103933  -4.235 2.29e-05 ***
ReasonLeakage and Burst  -0.299258   0.123353  -2.426 0.015265 *  
RegionMainland            0.806694   0.166710   4.839 1.31e-06 ***
months.since.spill24      0.369011   0.158085   2.334 0.019582 *  
months.since.spill36      0.993695   0.150844   6.588 4.47e-11 ***
months.since.spill48      0.780223   0.139148   5.607 2.06e-08 ***
months.since.spill6       0.664436   0.229903   2.890 0.003852 ** 
months.since.spill60      1.009377   0.259291   3.893 9.91e-05 ***
months.since.spillCLU     0.430998   0.212747   2.026 0.042778 *  
bareground.cover          0.003596   0.001370   2.624 0.008692 ** 
litter.cover             -0.006325   0.001808  -3.499 0.000467 ***
Distance.to.nearest.bush -0.003132   0.001457  -2.150 0.031570 *  
buildings                 0.092977   0.039433   2.358 0.018382 *  
roads                     0.152469   0.089965   1.695 0.090121 .  

现在的问题是那个月,因为泄漏是8级的一个因子。输出显示了6级。我了解拦截器已采取其中一个级别。但是,我无法解释最后一级的下落。拜托,有人可以帮我吗?谢谢。

0 个答案:

没有答案