如何在散点图中显示相关系数？答案

【问题标题】：How do I display a correlation coefficient in a scatterplot?如何在散点图中显示相关系数？
【发布时间】：2021-12-27 11:10:00
【问题描述】：

在散点图中，我想沿描述 x 和 y 之间关系的方程显示相关系数。我已经创建了我的数据材料，这是我目前的代码：

library(tidyverse)

# Creation of datamaterial

salary <- c(95, 100, 105, 110, 120, 124, 135, 150, 165, 175, 225, 230, 235, 260)
height <- c(160, 150, 182, 165, 172, 175, 183, 187, 174, 193, 201, 172, 180, 188)
fakenumbers <- data.frame(salary, height)

cor(height, salary, method = c("pearson"))

# Creation of scatterplot

r <- ggplot(fakenumbers, aes(x = height, y = salary)) + 
  geom_point(size = 3, shape = 21, color = "black", fill = "blue") + 
  labs(y = "Hourly salary 
       (sek)", x = "height (cm)", title = "Relationship between height and salary (made up data)") + 
  theme_classic() + theme(plot.title = element_text(hjust = 0.5, size = 18), 
                          axis.title = element_text(size = 15), 
                          axis.title.y = element_text(angle = 0, vjust = 0.5), 
                          axis.text = element_text(size = 11))

# Adding a regressionline

r + geom_smooth(method = lm, formula = y ~ x, se = FALSE)

在坐标系内部，在回归线旁边，我想显示一个“r = 0.588”和一些描述线性关系的方程。我怎样才能做到这一点，最好使用 ggplot() 或其他一些函数？

【问题讨论】：

标签： r ggplot2 linear-regression scatter-plot pearson-correlation

【解决方案1】：

我们可以使用ggpubr 包来实现，将stat_cor(p.accuracy = 0.001, r.accuracy = 0.01) 添加到您的代码中：

library(ggpubr)
library(tidyverse)

r <- ggplot(fakenumbers, aes(x = height, y = salary)) + 
  geom_point(size = 3, shape = 21, color = "black", fill = "blue") + 
  stat_cor(p.accuracy = 0.001, r.accuracy = 0.01)+
  labs(y = "Hourly salary 
       (sek)", x = "height (cm)", title = "Relationship between height and salary (made up data)") + 
  theme_classic() + theme(plot.title = element_text(hjust = 0.5, size = 18), 
                          axis.title = element_text(size = 15), 
                          axis.title.y = element_text(angle = 0, vjust = 0.5), 
                          axis.text = element_text(size = 11))

【讨论】：

【解决方案2】：

这里是基本的 R 方式。定义一个公式fo，计算回归，定义一个eqation。

corr <- cor(height, salary, method = c("pearson"))

fo <- salary ~ height
fit <- lm(fo, fakenumbers)
(eq <- paste0(all.vars(fo)[1], ' ~ ', paste0(round(coef(fit), 2),
              gsub('\\*\\(Intercept\\)', '', 
                   paste0('*', names(coef(fit)))), collapse=' + ')))
# [1] "salary ~ -281.58 + salary ~ 2.49*height"

然后在plot()、abline()和text()中使用变量。

plot(fo, fakenumbers, pch=20, col=4,
     xlab='height (cm)', ylab='Hourly salar (sec)',
     main='Relationship between height and salary (made up data)')
abline(fit, col=4)
text(149, 250, bquote(italic('r=')~.(round(corr, 3))), adj=0, cex=.8)
text(149, 235, eq, adj=0, cex=.8)

数据：

fakenumbers <- structure(list(salary = c(95, 100, 105, 110, 120, 124, 135, 150, 
165, 175, 225, 230, 235, 260), height = c(160, 150, 182, 165, 
172, 175, 183, 187, 174, 193, 201, 172, 180, 188)), class = "data.frame", row.names = c(NA, 
-14L))

【讨论】：

【解决方案3】：

另一种方式：

round(cor(height, salary, method = c("pearson")), 4) -> corr

然后使用geom_text显示相关系数：

r +
  geom_smooth(method = lm, formula = y ~ x, se = FALSE) +
  geom_text(x = 152, y = 250,
            label = paste0('r = ', corr),
            color = 'red')

【讨论】：