Question

假设，我有一些数据点定义了5个不同的图形，如下图所示。

如何在y轴的值上绘制这些的平均图表？
我不能直接这样做，因为不同图形的数据点在x轴上的值不同。

Answer 1

import numpy as np

def line_tuple(filename,cols=(0,1)):
    return np.loadtxt(filename,usecols=cols,unpack=True)

#parse each line from the datafile into a tuple of the form (xvals,yvals)
#store that tuple in a list.
data = [line_tuple(fname) for fname in ("line1.txt","line2.txt","line3.txt","line4.txt","line5.txt")]

#This is the minimum and maximum from all the datapoints.
xmin = min(line[0].min() for line in data)
xmax = max(line[0].max() for line in data)

#100 points evenly spaced along the x axis
x_points = np.linspace(xmin,xmax,100)

#interpolate your values to the evenly spaced points.
interpolated = [np.interp(x_points,d[0],d[1]) for d in data]

#Now do the averaging.
averages = [np.average(x) for x in zip(*interpolated)]

#put the average value along with it's x point into a file.
with open('outfile','w') as fout:
    for x,avg in zip(x_points,averages):
        fout.write('{0} {1}\n'.format(x,avg))

现在我绘制它：

plot 'line1.txt' w l, \
     'line2.txt' w l, \
     'line3.txt' w l, \
     'line4.txt' w l, \
     'line5.txt' w l, \
     'outfile' w l

绘制5幅图的“平均”图

1 个答案: