绘制组与组中变量的平均值

Question

I have a CSV file in the format of:

BUFFER_SIZE,RUN,DURATION
1000,1,0.5
1000,2,0.62
1000,3,0.48
1000,4,0.59
2000,1,0.44
2000,2,0.35
2000,3,0.29
2000,4,0.41
...

(the data is fake, it's just to illustrate my example)

I want to plot buffer_size vs mean(duration) .

I can group and compute means without a problem:

bench_results = pd.read_csv('bench_results.csv')
bench_by_size = bench_results.groupby('BUFFER_SIZE')
bench_by_size.mean()

which gives me the expected results.

plot(bench_results.groupby('BUFFER_SIZE').mean()['DURATION']) is almost what I want, except that I want the X-axis to be BUFFER_SIZE.

This is ugly but gives what I want:

Xvals = []
Yvals = []
for key, grp in bench_results.groupby(['BUFFER_SIZE']):
    Xvals.append(key)
    Yvals.append(mean(grp['DURATION']))
plot(Xvals, Yvals)

Is there a better way to do that? I would like to avoid itearing over the GroupBy object.

Answer 1

plt.plot(bench_by_size.mean()['DURATION']) should work. For example,

import pandas as pd
import matplotlib.pyplot as plt

bench_results = pd.DataFrame(
    {'BUFFER_SIZE': [1000, 1000, 1000, 1000, 2000, 2000, 2000, 2000],
     'DURATION': [0.5, 0.62, 0.48, 0.59, 0.44, 0.35, 0.29, 0.41],
     'RUN': [1, 2, 3, 4, 1, 2, 3, 4]})

# bench_results = pd.read_csv('data')
bench_by_size = bench_results.groupby('BUFFER_SIZE')
means = bench_by_size.mean()
plt.plot(means['DURATION'], linestyle='-', marker='o', markersize=10)
plt.xlabel(means.index.name)
plt.ylabel('DURATION')
plt.show()

yields