Pandas matplotlib graphing multiple subplots

Question

I am trying to create a graph that graphs event count on the y axis and time on the x axis. I want to have two different subplots, each representing a unique process id (to monitor how that process affects event_count). Here is some of my data:

  ProcessID      Time  event_count
479         1592      1.49        62760
480         1592      1.49       379620
481         1592      1.49       117124
482         1592      2.62       450024
483         1592      2.62       126941
484         1592      3.75       126360
485         1592      3.76       468223
486         1592      4.88       400239
487         1592      4.88       129450
488         1592      6.01       441982
489         1592      6.01       129858
490         1592      7.14        88848
491         1592      7.14       421015
492         1592      7.14       125487
493         1592      8.27       427974
494         1592      8.27       131260
495         1592      9.40       441375
496         1592      9.40       129779
497         1592     10.53       414021
498         1592     10.53       131006
499         1592     11.66       434822
500         1592     11.66       128453
501         1592     12.79        51726

 ProcessID      Time  event_count
52715       7908      1.49        95615
52716       7908      2.62        95974
52717       7908      3.75        95174
52718       7908      3.76       116662
52719       7908      4.88        74974
52720       7908      4.88       102559
52721       7908      6.01        74307
52722       7908      6.01       108027
52723       7908      7.14       110227
52724       7908      8.27        83922

This is what I have so far to get to this point:

df = pd.read_csv('Book3.csv')
df = df.rename(columns={'Time ': 'Time'})
df['Time'] = df.Time.astype(float)
df2 = df.ix[:,['ProcessID','Time', 'event_count']].query('ProcessID == 1592')
df3 = df.ix[:,['ProcessID','Time', 'event_count']].query('ProcessID == 7908')

Any ideas on how to achieve this using pandas matplotlib?

sacuL · Accepted Answer

You can use groupby and iterate through your groups. To show a rough example on your data:

import matplotlib.pyplot as plt

fig, axes = plt.subplots(2)
for subplot_number, (pID, data) in enumerate(df.groupby('ProcessID')):
    axes[subplot_number].plot(data['Time'], data['event_count'])
    axes[subplot_number].set_title('Process ID: {}'.format(pID))
    axes[subplot_number].set_ylabel('Event Count')
    axes[subplot_number].set_xlabel('Time')

plt.tight_layout()
plt.show()

Edit: To make both on the same plot (as opposed to subplots), you can use more simple syntax because you don't have to specify the ax, putting the whole thing in a list comprehension, something along the lines of:

[plt.plot(data.Time, data.event_count, label='Process ID: {}'.format(pID)) for pID, data in df.groupby('ProcessID')]

plt.ylabel('Event Count')
plt.xlabel('Time')

plt.legend(loc=1)
plt.tight_layout()
plt.show()

Or even more simple, use pandas built in plot (which builds on matplotlib):

fig, ax = plt.subplots()
df.set_index('Time').groupby('ProcessID')['event_count'].plot(ax=ax, legend=True)

plt.show()

Personally, I find this harder to customize

Pandas matplotlib graphing multiple subplots

Answers (1)

Related Questions