Averaging multiple graph lines into one, python & matplotlib

Question

I am brand new to coding, please be gentle! I have a main script that runs a simulation of particles colliding with each other and walls in micro-gravity conditions. This part of the script outputs individual data files containing: timestep, vtotal. There are 15 particles so I get out 15 txt files.

N_max = sim.getNumTimeSteps()
particleData = [ [] for x in range(len(sim.getParticleList()))]
for n in range (N_max):
    sim.runTimeStep()
    if (n%1000==0):
        particles = sim.getParticleList()
        for i in range(len(sim.getParticleList())):
            print i
            x, y, z = particles[i].getVelocity()
            particleData[i].append( (n, x, y, z ))
print len(sim.getParticleList())
            
for i in range(len(sim.getParticleList())):
    with open("{0:d}.dat".format(i), "w") as f:
        for j in particleData[i]:
            f.write("%f,%f \n" % (j[0], (math.sqrt(float(j[1])**2+float(j[2])**2+float(j[3])**2)) ))
sim.exit()

The end result I need to work toward is a graph of the mean of those 15 particles over time. For example, in this simulation it was running for 22000 timesteps, at increments of 1000. Correct me if I am wrong, but the mean should be (vtotal1+vtotal2+vtotal3+...vtotal15)/per increment. When that is plotted over time, a single line represents the mean velocity of the 15 particles from the simulation? Here is a version of what I was doing that was adapted from another averaging attempt.

import matplotlib
matplotlib.use('agg')
import matplotlib.pyplot as plt
import csv
import math
import numpy as np

x = []
y = []
y_mean = np.array([1 for _ in range(22000/1000)])
fig, axes = plt.subplots(nrows=1, ncols=1, figsize=(10, 7))
for i in range(15):
    x = []
    y = []
    with open("{}.dat".format(i),'r') as csvfile:
        plots = csv.reader(csvfile, delimiter=',')
        for row in plots:
            x.append(float(row[0]))
            y.append(float(row[1]))
            y_mean[int(float(row[0]) / 1000)] += y[-1]
        axes.plot(x,y, color='skyblue', label="Total v {}".format(i+1))
        axes.plot(x,y_mean, color='olive', label="Average v {}".format(i+1))
        plt.title('Particles Over Time')
        plt.xlabel('Timestep Number')
        plt.grid(alpha=.5,linestyle='--')
        plt.ylabel('Velocity')
        plt.xlim(0, 2000)
        plt.show()
        plt.autoscale(enable=True, axis=y, tight=True)
plt.legend()
plt.savefig("round2avgs.png")
y_mean = np.asarray(y) / 15

I just don't know what's going wrong. Any assistance is appreciated.

Here is what the plot looks like: (i.imgur.com/VknfAyu.png) Here is what the velocity one had looked like before: (i.imgur.com/4AZetMc.png) And what an attempt at averaging looked like before: (i.imgur.com/Ap4cqIq.png) — spokati
– spokati, Commented Jun 25, 2020 at 5:14

adrianp · Accepted Answer · 2020-06-25 10:23:52Z

1

Normally, you should split your data processing and visualization into two different steps.

Say you have a 5 CSVs, all having the same data:

Let's name this 1.dat, 2.dat ... 3.dat.

Import the libraries and load the data

import csv

import matplotlib.pyplot as plt
import numpy as np

x = []
ys = []
for i in range(5):
    with open(f'{i+1}.dat') as data_file:
        data = csv.reader(data_file, delimiter=',')
        y = []
        for row in data:
            if i == 0:
                x.append(float(row[0]))
            y.append(float(row[1]))
        ys.append(y)

Calculate the means per timestep using numpy

means_per_timestep = np.array(ys).mean(axis=0)

Plot it

plt.plot(x, means_per_timestep)

Is this what you were expecting?

answered Jun 25, 2020 at 10:23

adrianp

1,0191 gold badge9 silver badges14 bronze badges

Sign up to request clarification or add additional context in comments.

2 Comments

spokati Over a year ago

Yes! That looks more like it. Thank you!

adrianp Over a year ago

I'd appreciate if you hit the green tick next to my answer if you found it helpful :-)

Collectives™ on Stack Overflow

Averaging multiple graph lines into one, python & matplotlib

1 Answer 1

2 Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

2 Comments

Your Answer

Sign up or log in

Post as a guest

Related