时间戳

首页 > 解决方案 > Plotting time series by year only in Matplotlib

python - Plotting time series by year only in Matplotlib

问题描述

I have a series that have dates in datetime format that look like:

2015-05-10  00:00:00

I would like to plot two different values in my dataframe over the same time periods and I've tried:

plt.subplot(211)
plt.plot(df['Date'].dt.year, df['Avg'], color='r')
plt.show()
plt.subplot(212)
plt.plot(df['Date'].dt.year, df['MW'], color='b')
plt.show()

But this gives me a strange format that can't even be dealt with by changing the ['Date'] column outside of plotting.

It plots as:

2015.0    2015.5   2016.0   2016.5.....etc

Tried changing to float form as well but to no avail.

All I want is one yearly tick over the entire series from 2015 to 2018 but when I convert to a string or integer I get something that looks like:

enter image description here

标签： pythonpython-3.xpandasdatetimedataframe

解决方案

.set_xticks在我看来，这是设置您在 x 轴上看到的值的最易读的方法。

import pandas as pd
import matplotlib.pyplot as plt

data =[ {"Date":"2015-05-10", "Avg":34}, {"Date":"2016-05-10", "Avg":32}, 
        {"Date":"2017-05-10", "Avg":31}, {"Date":"2015-05-10", "Avg":31}, 
        {"Date":"2015-05-10", "Avg":26}, {"Date":"2015-05-10", "Avg":29}]

df = pd.DataFrame(data)

ax = plt.subplot(111)
ax.plot(df['Date'].dt.year, df['Avg'], color='r')
ax.set_xticks([2015, 2016, 2017])

plt.show()

看起来您也遇到了问题，因为您每年都有多个值，因此它绘制了您看到的垂直线。使用 pandas.groupby函数的聚合值似乎是您想要的。

df["Year"] = df["Date"].dt.year

df = df.groupby("Year").mean().reset_index()

ax = plt.subplot(211)
ax.plot(df.Year, df['Avg'], color='r')
ax.set_xticks([2015, 2016, 2017])

plt.show()

推荐阅读