Нарисуйте текст и фигуры на расстоянии x, y от полей графика в R

Question

Для визуализации мощности с различными значениями N, размерами эффектов и дисперсиями, я моделирую и отображаю данные ниже:

Библиотеки:

library(repr)
library(plotrix)

# N = sample size
N = 1000

# Sample 1
x_gi = rnorm(N, mean=38.3, sd=8)
# Plot
x_gi_hist=hist(x_gi ,
     breaks=25,
     main=expression(paste(italic(bar(x)[0]))),
     xlab="taurine per unit of something",
     col="firebrick")

# Sample 2
x_gf = rnorm(N, mean=20, sd=7)
# Plot
x_gf_hist=hist(x_gf ,
     breaks=25,
     main=expression(paste(italic(bar(x)[gf]))),
     xlab="taurine per unit of something",
     col="grey70")

# Both samples together
#####################
# H0: Grain-inclusive distribution
# Plot
plot(x_gi_hist,
     col= c("firebrick",rgb(0.8,0,0,1/2))[cut(x_gi_hist$breaks, c(-Inf, max(sort(x_gi)[(1: (length(x_gi)*0.05))]),Inf))],   
     xlim=c(0,75), ylim=c(0,N/5),
     main="", xlab="taurine per unit of something",
     cex.lab=1.5,
     border=c("firebrick",rgb(0.7,0,0,1/4)))
# H0 label
text(x=mean(x_gi), y=max(x_gi_hist$counts) + (N/15),
    label = expression(paste(italic(H[0]))),
    cex=2)
# Mean label 
text(x=mean(x_gi), y=max(x_gi_hist$counts) + (N/30),
    label = expression(paste(italic(bar(x)[0]))),
    cex=2)

# Critical value
ablineclip(v=max(sort(x_gi)[(1: (length(x_gi)*0.05))]+0.8),
           lwd=2, lty=2, col="red",
           y1=0, y2=max(x_gf_hist$counts)-5)
# Critical value
text(x=max(sort(x_gi)[(1: (length(x_gi)*0.05))]+0.8), y=max(x_gf_hist$counts)-4,
    label="critical value", adj=0,cex= 1.5)

# Effect size arrow
arrows(x0=mean(x_gf)+4, x1=mean(x_gi)-4,
       y0=max(x_gi_hist$counts) + (N/30), y1=max(x_gi_hist$counts) + (N/30),
       lty=1, code=3, lwd=3, col="black")
# Effect size label
text(x=mean(x_gf) + (1/4 * mean(x_gi)), y=max(x_gi_hist$counts) + (N/25),
    label = expression(paste(italic(d))),
    cex=2)

#####################
# HA: Grain-free distribution
# Plot GF
plot(x_gf_hist,
     col= c(rgb(0,0,0,0.2), rgb(0,0,0,0.5))[cut(x_gf_hist$breaks, c(-Inf, max(sort(x_gi)[(1: (length(x_gi)*0.05))]),Inf))],  
     xlim=c(0,75),ylim=c(0,N/5),
     add=T, border=rgb(0,0,0,0.2))
# HA label
text(x=mean(x_gf), y=max(x_gi_hist$counts) + (N/15),
    label = expression(paste(italic(H[A]))),
    cex=2)
# Mean label 
text(x=mean(x_gf), y=max(x_gi_hist$counts) + (N/30),
    label = expression(paste(italic(bar(x)[gf]))),
    cex=2, font=2)

Это работает, когда N > 100, но ломается, когда N <10. </p>

Вопрос : есть ли способ генерировать разумные графики для более низких значений N, скажем, когда N = 10

Здесь Вот пример:

Спасибо за ваше время!