pandas 图中错误栏的样式 [英] Style of error bar in pandas plot
问题描述
我想用以下样式绘制带有错误栏的折线图.
I'd like to plot line chart with error bar with the following style.
但是，熊猫图只用垂直线绘制误差线.
However, pandas plot draws error bars with only vertical line.
pd.DataFrame([1,2,3]).plot(yerr=[0.3,.3,.3])
如何更改熊猫图的误差线样式?
How do I change style of error bar for pandas plot?
版本为:
- pandas'0.18.0'
- matplotlib'1.5.1'
一个原因似乎是使用了seaborn风格.下面的代码给出了漂亮的样式图.
One of the reason seems using the seaborn style. The following code give the nice style plot.
# plt.style.use('seaborn-paper')
pd.DataFrame([1,2,3]).plot(yerr=[0.3,.3,.3],capsize=4)
但是，我有理由继续使用淡淡的风格...请帮忙.
But, I have a reason to keep using seaborn style... Please help.
推荐答案
当您在DataFrame上调用plot时，可以使用capsize kwarg(将其传递给 plt.errorbar ):
You can change the capsize inline when you call plot on your DataFrame, using the capsize kwarg (which gets passed on to plt.errorbar):
pd.DataFrame([1,2,3]).plot(yerr=[0.3,.3,.3],capsize=4)
或者，您可以使用 rcParams 更改此设置.
Alternatively, you can change this setting using rcParams
您可以通过打印plt.rcParams['errorbar.capsize']来查找默认的错误栏上限大小.如果该值为0(这就是我怀疑您目前没有错误栏上限的原因)，则可以使用以下方法将错误栏上限的默认大小设置为非零值:
You can find out what your default errorbar cap size is by printing plt.rcParams['errorbar.capsize']. If that is 0 (which is why I suspect you are currently getting no errorbar caps), you can set the default size of the errorbar caps to something nonzero, using:
plt.rcParams['errorbar.capsize']=4
请确保在任何绘图脚本的开头都有该内容.
Make sure to have that at the beginning of any plotting script.
似乎使用seaborn-paper样式将瓶盖厚度设置为0.您可以使用capthick kwarg覆盖它:
It seems using the seaborn-paper style sets the cap thickness to 0. You can override this with the capthick kwarg:
plt.style.use('seaborn-paper')
pd.DataFrame([1,2,3]).plot(yerr=[0.3,.3,.3],capsize=4,capthick=1)
这篇关于 pandas 图中错误栏的样式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！
