可视化逻辑回归 GAM
R 中的广义可加模型（GAM）非线性建模
Noam Ross
Senior Research Scientist, EcoHealth Alliance
变换刻度
对数几率图
plot(binom_mod)
转换偏效应
plot(binom_mod, pages = 1, trans = plogis)
转换偏效应 (2)
plot(binom_mod, pages = 1, trans = plogis)
添加截距
plot(binom_mod, pages = 1, trans = plogis,
shift = coef(binom_mod)[1])
添加截距 (2)
plot(binom_mod, pages = 1, trans = plogis,
shift = coef(binom_mod)[1])
纳入截距的不确定性
plot(binom_mod, pages = 1, trans = plogis,
shift = coef(binom_mod)[1],
seWithMean = TRUE)
改进绘图
plot(binom_mod, pages = 1, trans = plogis, shift = coef(binom_mod)[1],
seWithMean = TRUE, rug = FALSE, shade = TRUE,
shade.col = "lightgreen", col = "purple")
