如何表示线性回归模型的结果？

Question

使用面板数据，我在人口变量变量age上回归了144列（时间步长）

我将统计信息保存在名为Age的数据框中

dput(head(Age, 10))
structure(list(term = c("(Intercept)", "act1_1", "(Intercept)", 
"act1_2", "(Intercept)", "act1_3", "(Intercept)", "act1_4", "(Intercept)", 
"act1_5"), estimate = c(44.3846208128138, -0.000384107652327793, 
44.3174018132914, -5.98548832764511e-05, 44.2450887508948, 0.000196982881599239, 
44.2072004527491, 0.000308764204242224, 44.181318653263, 0.000390867379854134
), std.error = c(0.168699503373036, 0.000204411891544783, 0.168567124457874, 
0.000184836499264514, 0.16808328410892, 0.000136569949465428, 
0.168158343129598, 0.000131172551491056, 0.168235711394923, 0.000132127095441471
), statistic = c(263.098704651599, -1.8790866295743, 262.906554026, 
-0.32382610314857, 263.233128656764, 1.44235889645037, 262.890318910189, 
2.35387816073149, 262.615578386625, 2.95826816254565), p.value = c(0, 
0.0602502440977185, 0, 0.746073789321371, 0, 0.14922009840405, 
0, 0.0185903375815117, 0, 0.00309810897731306)), row.names = c(NA, 
-10L), class = c("tbl_df", "tbl", "data.frame"))

如何以有意义的方式绘制p值（考虑到act1_1时间步长）？

谢谢

Answer 1

library(ggplot2)
library(dplyr)
df <- df %>% 
  dplyr::mutate(term = c("Interecpt1", "act1_1", 
                         "Intercept2", "act1_2",
                         "Intercept3", "act1_3",
                         "Intercept4", "act1_4",
                         "Intercept5", "act1_5"),
                signif = ifelse(p.value > 0.05 , 1, 0),
                signif = as.factor(signif)) # alpha = 5%
ggplot2::ggplot(df) +
  geom_point(aes(x = term, y = p.value, col = signif), size = 2, show.legend = F) +
  geom_hline(yintercept = 0.05, linetype = 2, col = "red") +
  coord_flip() +
  scale_color_manual(values = c("darkblue", "darkred")) +
  theme_bw()

Answer 2

假设x轴是日期列，则可以使用函数scale_x_date()并设置需要缩放的值，x轴看起来会好很多。

示例语句如下所示。

p1 <- p1 + scale_x_date(date_breaks = "1 week")

• 由于未提供实际的ggplot代码，因此我不确定您的x轴是什么。