I want to visualize a line plot of time series with different classes in the background in python.
Suppose if I have time series of 5 minutes with 3 classes as shown in the table 1.
| Time start | Time End | Class |
|---|---|---|
| 00:00:00 | 00:00:30 | 1 |
| 00:00:31 | 00:01:30 | 2 |
| 00:01:31 | 00:04:00 | 3 |
| 00:04:01 | 00:05:00 | 2 |
I also have pressure values every 10 seconds shown in table 2.
| Time End | Pressure |
|---|---|
| 00:00:10 | 0.2 |
| 00:00:20 | 0.3 |
| 00:00:30 | 0.53 |
| . | . |
| . | . |
| 00:04:50 | 0.7 |
| 00:05:00 | 0.92 |
I want to plot a line graph on top of the classes with different colors as shown in the figure below. The line plot follows the table 2 and the background color of the line plot follows table 1.
I am open to use any python library that could help me complete this visualization.
CodePudding user response：
You can use the following snippet:
# convert timestamp to seconds
df2['Time End'] = pd.to_timedelta(df2['Time End']).dt.total_seconds()
df1['Time start'] = pd.to_timedelta(df1['Time start']).dt.total_seconds()
df1['Time End'] = pd.to_timedelta(df1['Time End']).dt.total_seconds()
# plot line
ax = df2.plot(x='Time End', y='Pressure')
ax.set_ylabel('Pressure')
# plot spans
colors = {1: '#D9E8FB', 2: '#F7CECC', 3: '#FFF2CD'}
for _, (x1, x2, c) in df1.iterrows():
ax.axvspan(x1, x2, color=colors[c])
