在matplotlib中显式添加x轴时，图形会变形

Question

I'm preparing some data for analysis. The data contains timestamped pollution measurements from different sensors. To get a feeling for the data and make sure I get the data correctly and that there aren't big gaps in it, I wanted to plot not only the pollution, but the pollution against the timestamp. I expect an almost identical plot like the one plotting only the pollution against the data index, but with small gaps for the periods when the sensors didn't work. Unfortunately something goes wrong and the two plots are completely different. The problem evidently lies in my approach of plotting the second plot. The data point are somehow connected back and fourth in time, the whole thing makes no sense. How do I get this to work?

This is the code producing the plots.

import matplotlib.pyplot as plt
import pandas

#load data set from csv
df = pandas.read_csv('csv_exports/measurements1207.csv')
#add column names to laoded data
df.columns = ['id', 'air_quality_index', 'from_date', 'latitude', 'longitude', 'pm_1',
              'pm_10', 'pm_25', 'pollution_level', 'sensor_id', 'source', 'till_date',
              'wind_deg', 'wind_speed']
df_pm = df[['sensor_id', 'from_date', 'pm_10', 'pm_25']]

#last 20 thousend pm10 and pm25 values
last_20k = df_pm.tail(20000)
print("first row example")
print(last_20k.iloc[0])
first_row_sensor_id = last_20k.iloc[0].loc['sensor_id']

one_sensor = df_pm.loc[df['sensor_id'] == first_row_sensor_id][['from_date', 'pm_10', 'pm_25']]
one_sensor = one_sensor.dropna()
one_sensor[['from_date']] = pandas.to_datetime(one_sensor['from_date'], format='%Y-%m-%d %H:%M:%S.%f')
#plot values
plt.plot(one_sensor[['pm_10', 'pm_25']])
plt.show() // first plot
plt.plot(one_sensor[['from_date']], one_sensor[['pm_10', 'pm_25']])
plt.show() // second faulty plot

Answer 1

As stated in the comment by @ImportanceOfBeingErnest the data needs to be sorted. Simply add a line like this to your code.

one_sensor = one_sensor.sort_values(['from_date'], ascending=[True])