In [5]:

    

%pylab inline

numbers = np.random.normal(0,1,100000000)


    

    




Populating the interactive namespace from numpy and matplotlib

In [11]:

    

def get_ecdf(data):
    x_sorted = np.sort(data)
    n = x_sorted.size
    y = np.arange(1, n+1) / n
    return x_sorted, y


    

In [12]:

    

x, y = get_ecdf(numbers)


    

In [13]:

    

plt.scatter(x=x, y=y)


    

    
Out[13]:





<matplotlib.collections.PathCollection at 0x2b2e57d5f9b0>






    

In [20]:

    

sampled_index = np.random.choice(np.arange(x.size), size=100)
plt.scatter(x=x[sampled_index], y=y[sampled_index])


    

    
Out[20]:





<matplotlib.collections.PathCollection at 0x2b2e57ac0cc0>






    

In [16]:

    

len(sampled_points)


    

    
Out[16]:





1000

In [ ]: