用ggplot躲避点和误差线 [英] Dodging points and error bars with ggplot

问题描述

考虑这些数据（注意 foo 实际上是一个因素。）：

  foo bar结果ci 
 1 a 0.683333333 0.247447165 
 2 b 0.941666667 0.180356565 
 3 c 0.783333333 0.335337789 
 1 d 0.866666667 0.204453706 
 2 e 0.45 0.303059647 
 3 f 0.325 0.340780173 
  

我想绘制多个 bar foo 值，其中结果和带有CI的错误栏。这是我做的：

  ggplot（ex，aes（foo，outcome，label = bar））+ 
 geom_point （位置= position_dodge（.1））+ 
 geom_errorbar（aes（ymin =结果 -  ci，ymax =结果+ ci），position = position_dodge（.1））+ 
 geom_text（hjust = 2） 
  

我得到：

但我希望它避开误差线和点我可以看到重叠。 使用 position_jitter 做到了，但它完全是随机的（或笨重） - 我不希望这样。

如何抵消个人观察？

或者这是一个错误与ggplot？ 此处的示例也显示此错误。

一种可能性是'bar'对 group >。请注意，我也闪避 geom_text 。

<$ p $ （数据= df，aes（x = foo，y =结果，组=条，标签=条））。 +
geom_point（position = dodge）+
geom_errorbar（aes（ymin =结果 - ci，ymax =结果+ ci），position = dodge）+
geom_text（hjust = 2，position =闪避）

Consider this data (note that foo is actually a factor.):

foo bar outcome ci
1   a   0.683333333 0.247447165
2   b   0.941666667 0.180356565
3   c   0.783333333 0.335337789
1   d   0.866666667 0.204453706
2   e   0.45    0.303059647
3   f   0.325   0.340780173

I want to plot multiple bars per foo value, with their outcome and error bars with CI. Here's what I do:

ggplot(ex, aes(foo, outcome, label = bar)) + 
  geom_point(position = position_dodge(.1)) + 
  geom_errorbar(aes(ymin = outcome - ci, ymax = outcome + ci), position = position_dodge(.1)) + 
  geom_text(hjust = 2)

I get:

But I wanted it to dodge the error bars and points so I can see the overlap. Using position_jitter did that, but it was totally random (or "clunky") — I don't want that.

How can I offset the individual observations?

Or is this a bug with ggplot? The example here also shows this error.

解决方案

One possibility is to group by 'bar'. Note that I also dodge the geom_text.

dodge <- position_dodge(.1)

ggplot(data = df, aes(x = foo, y = outcome, group = bar, label = bar)) + 
  geom_point(position = dodge) + 
  geom_errorbar(aes(ymin = outcome - ci, ymax = outcome + ci), position = dodge) + 
  geom_text(hjust = 2, position = dodge)

这篇关于用ggplot躲避点和误差线的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持IT屋！