python-如何使用重复的x轴绘制数据

Question

I would like to plot day dependent data over many years where years should be on the x-axis (2016,2017,2018 for example). What is a good approach to do this?

For every year I have a list of days which I would plot on the x-axis but of course python keeps this axis and plots all the data of different years above each other.

Any suggestions?

Code:

A shortened version of my dictionary L_B_1_mean looks like this:

2016018 5.68701407589
2016002 4.72437644462
2017018 3.39389424822
2018034 7.01093439059
2018002 8.79958946488
2017002 3.55897852367

the code:

data_plot = {"x":[], "y":[], "label":[]}
for label, coord in L_B_1_mean.items():
    data_plot["x"].append(int(label[-3:]))             
    data_plot["y"].append(coord)
    data_plot["label"].append(label)


# add labels
for label, x, y in zip(data_plot["label"], data_plot["x"], data_plot["y"]):
    axes[1].annotate(label, xy = (x, y+0.02), ha= "left")


# 1 channel different years Plot
plt_data = axes[1].scatter(data_plot["x"], data_plot["y"])

I construct my x-values here: data_plot["x"].append(int(label[-3:])) where I read the name tag like: 2016002 and get only the day value: 002

In the end I have 365 days per year and now i would like to plot the data of 2016 then 2017 and then 2018 all after each other instead of on top of each other

Answer 1

You have a dict

L_B_1_mean 

{'2016018': 5.68701407589,
 '2016002': 4.72437644462,
 '2017018': 3.39389424822,
 '2018034': 7.010934390589999,
 '2018002': 8.79958946488,
 '2017002': 3.55897852367}

plot using pandas:

import pandas as pd

You can simply create a pandas series from this dict:

s = pd.Series(L_B_1_mean)

2016018    5.687014
2016002    4.724376
2017018    3.393894
2018034    7.010934
2018002    8.799589
2017002    3.558979
dtype: float64

...and cast the strings in the index to dates:

s.index = pd.to_datetime(s.index, format='%Y%j')

2016-01-18    5.687014
2016-01-02    4.724376
2017-01-18    3.393894
2018-02-03    7.010934
2018-01-02    8.799589
2017-01-02    3.558979
dtype: float64

Then you can plot your data easily:

s.plot(marker='o')

plot using datetime and matplotlib:

import datetime as DT
import matplotlib.pyplot as plt

t = [DT.datetime.strptime(k, '%Y%j') for k in L_B_1_mean.keys()]
v = list(L_B_1_mean.values())

v = sorted(v, key=lambda x: t[v.index(x)])
t = sorted(t)

plt.plot(t, v, 'b-o')