Using SubPlots With LineCollection

Question

I am trying to plot a random walk constrained to move about a lattice.

To implement this constraint I am using hstack to format the segments for LineCollection from the matplotlib module.

I want four random walks to start in four quadrants all on the same plot. As my code stands now, I get four individual plots.

How do I specify ploting all the data on the same plot? #multiple 2D random walks

    from matplotlib import collections  as mc
    import numpy as np
    import pylab as plt

    steps = 1000
    coar = np.empty([steps,2],int)
    #random walk start cooridiates
    n1=np.array([50,50],int)
    n2=np.array([-50,50],int)
    n3=np.array([-50,-50],int)
    n4=np.array([50,-50],int)
    na = [n1,n2,n3,n4]
    #colors of the four random walks
    clr = ['g','c','m','y']

    with open("raw_ran_576001.txt","r") as rf:
        for j in range(len(na)): 
            for t in range(0,steps):
                bin=rf.read(2)      #reads two bits to generate random step of walk
                if(bin=="00"):
                    na[j][0]+=1
                elif(bin=="11"):
                    na[j][0]-=1
                elif(bin=="01"):
                    na[j][1]+=1
                elif(bin=="10"):
                    na[j][1]-=1
                coar[t] = na[j] 
            coart = coar.reshape(-1,1,2)
            segments = np.hstack([coart[:-1],coart[1:]])
             # figure out how to add different random walks in different colors
             #to same plot
            coll = mc.LineCollection(segments,color=clr[j])
            fig, ax=plt.subplots()          #just a figure and one subplot
            ax.set_axis_bgcolor('black')
            ax.add_collection(coll)         #this must be where points are ploted
            ax.autoscale()
            t=0
    plt.show()

What am I overlooking

btw I am using random bits generated from a radioisotope hardware random number generator.

score 0 · Accepted Answer · answered Jul 29 '15 at 07:56

Fun problem. It was actually super simple - you just had to take the fig, ax=plt.subplots() command outside your loop.

Your example didn't work for me because I don't have a copy of the file you're drawing from so I used numpy's random module to emulate it. I also used the plot command because it seemed a simpler way to do what you're looking for.

# Import what you need
import numpy as np
import pylab as plt

# Set the number of steps you're going to walk along
steps = 1000

# Set the random walk start coordinates
# for each of four random walks
n1=np.array([50,50],int)
n2=np.array([-50,50],int)
n3=np.array([-50,-50],int)
n4=np.array([50,-50],int)
na_list = [n1,n2,n3,n4]

# Set the colors of the four random walks
clr_list = ['g','c','m','y']

# Create one figure with one subplot
fig, ax=plt.subplots()
# Set the background color to black
ax.set_axis_bgcolor('black')

# Loop through the different random walks
for na, clr in zip(na_list, clr_list):

    # Create a list of coordinates
    # initiated by the start coordinates
    coar = np.ones([steps+1,2],int) * na

    # For each step figure out if you're 
    # going to walk right, left, up or down        
    for t in range(0,steps):

        # Set coar for the point after
        # this step (t+1) to be the point the
        # step starts at (t)
        coar[t+1] = coar[t]

        # Get a random number
        bin = np.random.randint(4)

        if(bin==0):
            # Step to the right (inc x by 1) 
            coar[t+1][0] = coar[t,0] + 1
        elif(bin==1):
            # Step to the left (dec x by 1) 
            coar[t+1][0] = coar[t,0] - 1
        elif(bin==2):
            # Step up (inc y by 1) 
            coar[t+1][1] = coar[t,1] + 1
        elif(bin==3):
            # Step down (dec y by 1) 
            coar[t+1][1] = coar[t,1] - 1

    # Plot these coordinates
    ax.plot(coar.T[0], coar.T[1], c=clr)

    # And show the starting point with a white triangle
    # just to make it clear where you started
    ax.scatter(coar[0,0], coar[0,1], marker='^', c='w', edgecolor='w', s=70, zorder=3)

# Autoscale the axis
ax.autoscale()
# And show the plot
plt.show()

Brilliant! Interesting fact about a random walk constrained like this: each and every point on the 2D lattice has a probability of unity of being reached as the number of steps approaches infinity. You are a beautiful godess – Dexileos 6 hours ago — Dexileos, Jul 30 '15 at 07:30

Using SubPlots With LineCollection

1 Answers1