Home / Questions / 361210
Accepted
jfmejiat
Asked: 2020-06-01 23:23:28 +0800 CST 2020-06-01 23:23:28 +0800 CST

Plot two ´labels' for the same axis in Matplotlib

  • 772

I am performing a data analysis from a Dataframe and I need to make a climogram, which is a graph in which the months of the year are located on the horizontal axis, while on the vertical axis they are represented on the left axis the average monthly temperatures by means of points joined by a line, while on the right axis, and by means of bars, the average monthly rainfall is represented. How can I do to generate the two 'labels' of temperature and precipitation on the y axis?

Below is an example of what a climogram looks like and the data of the Dataframe that I am using:

climogram

precipitación_y_temperatura_promedio_Medellín_5=pd.merge(precipitación_y_temperatura_promedio_Medellín_3.reset_index(), precipitación_y_temperatura_promedio_Medellín_4.reset_index())
precipitación_y_temperatura_promedio_Medellín_5.set_index('Mes',inplace=True)
print(precipitación_y_temperatura_promedio_Medellín_5)

             Precipitación Promedio (mm)  Temperatura Promedio (°C)
Mes                                                               
Enero                              63.2                       22.4
Febrero                            81.4                       22.7
Marzo                             129.1                       22.7
Abril                             170.7                       22.4
Mayo                              213.5                       22.6
Junio                             149.4                       22.9
Julio                             133.1                       23.1
Agosto                            139.7                       23.1
Septiembre                        181.8                       22.4
Octubre                           226.7                       21.8
Noviembre                         158.9                       21.8
Diciembre                         104.8                       21.9
python

1 Answers

  1. Best Answer

    The data to reproduce the example:

    import io

import matplotlib.pyplot as plt
import pandas as pd

data = io.StringIO("""\
Mes,Precipitación Promedio(mm),Temperatura Promedio (°C)
Enero,63.2,22.4
Febrero,81.4,22.7
Marzo,129.1,22.7
Abril,170.7,22.4
Mayo,213.5,22.6
Junio,149.4,22.9
Julio,133.1,23.1
Agosto,139.7,23.1
Septiembre,181.8,22.4
Octubre,226.7,21.8
Noviembre,158.9,21.8
Diciembre,104.8,21.9
""")

df = pd.read_csv(data, index_col="Mes")

    You can't use the same axis because by definition they would share marks, scale and axis labels Y, which we don't want. The key is to use matplotlib.axes.Axes.twinx to generate another instance of Axesbut sharing the axis Xof the old one.

    Then each is used for one of the graphs. Since the axis generated by twinxoverlaps the previous one, it must be used for the temperatures, changing the labels and marks of both Axes to the opposite side each one. In this way we avoid resorting to modifying the transparency that always looks bad.

    fig = plt.figure()
ax1 = fig.add_subplot(111)
ax2 = ax1.twinx()

df["Temperatura Promedio (°C)"].plot(color="red", ax=ax2, lw=2)
df["Precipitación Promedio(mm)"].plot(kind='bar', width=0.9,  ax=ax1)

ax2.set_ylim(0,  40)
ax1.yaxis.set_label_position("right")
ax1.yaxis.tick_right()
ax2.yaxis.set_label_position("left")
ax2.yaxis.tick_left()
ax1.spines['top'].set_visible(False)
ax2.spines['top'].set_visible(False)

fig.suptitle("Medellin")
ax2.set_ylabel("Temperatura Promedio (°C)")
ax1.set_ylabel("Precipitación Promedio (mm)")

plt.show()

    enter image description here

    • 1