example_runcounting.py

import numpy as np
import countoscope

import matplotlib.pyplot as plt # for plotting purposes
from matplotlib import cm

if __name__ == '__main__':
    Lx = 100.0 # box size x-dir 
    Ly = 100.0 # box size y-dir
    box_sizes = [16.0, 8.0, 4.0, 2.0, 1.0] # array of box sizes to probe

    a = 1.0 #radius of particles
    sep_sizes = [2*a,2*a, 2*a, 2*a, 4*a] #separation between boxes
    # conservative separation is to keep about a particle radius in between boxes
    # for smaller boxes if you don't want the code to run too long it's good to keep the separation large
    
    
    # load data
    filename = 'example_dataset'
    folder = 'test_data/'
    extension = '.txt'

    data = np.fromfile(f'{folder}{filename}{extension}', dtype=float, sep=' ') # load data array as a single big line
    data = data.reshape((-1, 3)) # reshape it so that it has x / y / t on each row

    # run the main counting code
    results = countoscope.calculate_nmsd(data=data, window_size_x=Lx, window_size_y=Ly, box_sizes=box_sizes, sep_sizes=sep_sizes)
    
    # load the results and save the data
    N2_mean = results.nmsd
    N2_std = results.nmsd_std
    N_mean = results.N_mean   
    N_var_mod = results.N_var_mod
    np.savez(f'{folder}{filename}_counted.npz', N2_mean=N2_mean, N2_std=N2_std, N_mean= N_mean, N_var_mod = N_var_mod, box_sizes=box_sizes, sep_sizes=sep_sizes)

    
   
    # a few plots for sanity checks 
    
    cmap = cm.viridis #pick a great colormap
    num_Boxes = len(box_sizes) #number of boxes 
    dt = 1.0 #time step in between frames
    nt = len(N2_mean[0][:])

    # basic plot
    for i, L in enumerate(box_sizes):
        plt.plot([it*dt for it in range(1, nt)], N2_mean[i, 1:], color=cmap(i/num_Boxes), label=f'Box Size = {L}')
        # the point at t=0 messes up a loglog graph so we don't plot it

    ax = plt.gca()
    ax.set_xlabel('Time $t$')
    ax.set_ylabel(r'Number fluctuations $\langle (N(t) - N(0))^2 \rangle$')
    ax.loglog()
    plt.legend()
    plt.savefig('test_data/number_fluctuations.png')
    plt.show()
    plt.clf()

    # make also rescaled plot
    for i, L in enumerate(box_sizes):
        plt.plot([it*dt/L**2 for it in range(1, nt)], [n2/L**2 for n2 in N2_mean[i, 1:]], color=cmap(i/num_Boxes), label=f'Box Size = {L}')
        # the point at t=0 messes up a loglog graph so we don't plot it

    ax = plt.gca()
    ax.set_xlabel('Rescaled time $t/L^2$')
    ax.set_ylabel(r'Rescaled Number fluctuations $\langle (N(t) - N(0))^2 \rangle / L^2$')
    ax.loglog()
    plt.legend()
    plt.savefig('test_data/number_fluctuations_rescaled.png')
    plt.show()