#!/usr/bin/env python
# -*- coding: utf-8 -*-

from tkinter import *
from random import choice     # http://docs.python.org/library/random.html
import numpy as np
import matplotlib.pyplot as plt     # http://matplotlib.sourceforge.net/api/pyplot_api.html#module-matplotlib.pyplot
import matplotlib.mlab as mlab    # http://matplotlib.sourceforge.net/api/mlab_api.html#module-matplotlib.mlab

window = Tk()
sizex = 400
sizey = 400
canvas = Canvas(window, width = sizex, height = sizey)
canvas.pack()
x = 200		# initial left-most edge of first ball
y = 1		# initial top-most edge of first ball
r = 4                  # ball diameter 
depx = 1            # displacement at each move in x direction
depy = 0

# create balls:
no_particles = 1600
dy = (sizey-2.)/(no_particles+1)         # y initial separation between balls
print(dy)
ball_list = []
for i in range(no_particles):
    ball = canvas.create_oval(x,y,x+r,y+r,fill="blue")
    y = y+dy
    ball_list.append(ball)

#moves
no_moves = 200
for j in range(no_moves):
    for ball in ball_list:
        canvas.move(ball, choice([-1,1])*depx, depy)
    canvas.after(1)
    canvas.update()

#analysis - histogram
# see http://matplotlib.sourceforge.net/examples/api/histogram_demo.html
xpos = []
for ball in ball_list:
    posi = canvas.coords(ball)
    xpos.append((posi[0]-x)/depx)
xh = np.array(xpos)  # see http://www.scipy.org/Cookbook/BuildingArrays
#  compute the mean mu and sigma  from xh (and/or theoretical value from random walk result)
mu = np.mean(xh)
sigma = np.std(xh)
fig = plt.figure()
ax = fig.add_subplot(111)
# print xh 
n, bins, patches = ax.hist(xh, 10, facecolor='green', alpha=0.75)
print(n,bins, patches)
# hist uses np.histogram to create 'n' and 'bins'. cf. http://docs.scipy.org/doc/numpy/reference/generated/numpy.histogram.html

ax.set_xlabel('X positions')
ax.set_ylabel('Occurences')

ax.grid(True)

plt.show()

#window.mainloop()