Notes on multiple random effects in lmer vs lme #Refered to this (also this), but Multiple random intercepts cannot be included in lme, only works with lmer m1<-lmer(AGBGI~logtotal_BA+lognbsp+logSA+ (1|Eco)+(1|PLOT_ID), weights=wi, can) summary(m1)$co m2<-lme(AGBGI~logtotal_BA+lognbsp+logSA, random=list(Eco=~1, PLOT_ID=~1), weights=~1/wi, can)#this only produces the same coefs when modelled with only PLOT_ID summary(m2)$tT #but Note that "random=list(Eco=~1)" in lme produces indentical coefs with "(1|Eco)" in lmer
#Nested random intercept (plot id nested to eco) works with lme, same coefs with lmer m1<-lmer(AGBGI~logtotal_BA+lognbsp+logSA+ (1|Eco/PLOT_ID), weights=wi, can) summary(m1)$co m2<-lme(AGBGI~logtotal_BA+lognbsp+logSA, random=~1|Eco/PLOT_ID, weights=~1/wi, can) summary(m2)$tT
geom_bar, 正負の棒グラフごとにアスタリスクを付ける geom_text(aes(y = logBAI_AI_mean + 0.005*sign(logBAI_AI_mean), label = "***"), position = position_dodge(width = 0.9), size = 8 , angle = 0)+ ※ 上記ではカテゴリごとに分けられないので、その場合は: annotate("text", x = 1, y =0.018, label = expression(paste("***")), parse=TRUE, size=6)+ annotate("text", x = 2, y = -0.008, label = expression(paste("*")), parse=TRUE, size=6)+