Pandas plot stacked and unstacked bar-CodePudding

I am new to pandas. I am trying to visualize the time my program takes for different tasks. I read my logs and calculated the average time, and it now looks like this:

As I have only one row, I need to draw these tree time values in a bar plot. Time1 should be a single bar and Time2 and Time3 being combined. Time3 is supposed to be on top of Time2 with a different color. What i did so far is:

axis = time_pd.plot(kind='bar', ylabel="Time [sek]", xlabel="Phase", figsize=(11,11), rot=0)
legend = axis.legend(["time1", "time2", "time3"], fontsize=26, loc='center left', bbox_to_anchor=(1.0, 0.93))

The result is 3 bars on only one label on the x-axis:

Can someone give me any advice how to set up 2 Bars with Time1 and Time2 with Time3 being on top of Time2?

CodePudding user response：

The solution is to plot to figures and merge the result.

import pandas as pd
import matplotlib.pyplot as plt

df = pd.DataFrame({'time1':[3], 'time2':[2], 'time3':[1]})

ax = df['time1'].plot.bar(ylabel="Time [sek]", xlabel="Phase", figsize=(8,8), color='green', position=1)
df[['time2', 'time3']].plot.bar(stacked=True, sharex=True, ax=ax, position=0)
legend = ax.legend(["time1", "time2", "time3"], fontsize=12, loc='center left', bbox_to_anchor=(1.0, 0.93))
plt.xlim([-1, 1])
plt.show()

Output

Comment

I use pandas.plot twice, the first time "plain" and the second time "stacked". This normally produces two figures. To have only one, I re-use the axis by ax=ax in the second call.

To avoid that both plots are on each other and the first can not be seen, I use the position keyword. The limits for the x-axis have to be set, too, or some parts of the plot are outside of the visual area.