In [44]:

    

using PyPlot
include("utils.jl")
using DiffusionTutorialUtilities
utils = DiffusionTutorialUtilities
A,xy = utils.load_graph_and_coords("newman-netsci")
utils.cgraphplot(A,xy,filled=false)


    

    




WARNING: replacing module DiffusionTutorialUtilities






    
Out[44]:

In [34]:

    

include("diffusions.jl")
using DiffusionAlgorithms
P = utils.normout!(A)


    

    




WARNING: replacing module DiffusionAlgorithms






    
Out[34]:





379x379 sparse matrix with 1828 Float64 entries:
	[2  ,   1]  =  0.5
	[3  ,   1]  =  0.5
	[4  ,   1]  =  0.0294118
	[5  ,   1]  =  0.037037
	[16 ,   1]  =  0.047619
	[44 ,   1]  =  0.25
	[113,   1]  =  0.0666667
	[131,   1]  =  0.0833333
	[250,   1]  =  0.166667
	[259,   1]  =  0.333333
	⋮
	[375, 376]  =  0.333333
	[377, 376]  =  0.333333
	[207, 377]  =  0.2
	[375, 377]  =  0.333333
	[376, 377]  =  0.333333
	[70 , 378]  =  0.0555556
	[303, 378]  =  0.0833333
	[379, 378]  =  0.333333
	[70 , 379]  =  0.0555556
	[303, 379]  =  0.0833333
	[378, 379]  =  0.333333

In [35]:

    

function single_seed_pagerank(P,alpha,seed)
    n = size(P,1)
    @assert 1 <= seed <= n
    v = zeros(size(P,1))
    v[seed] = (1.-alpha)
    x = (eye(n) - alpha*P') \ v
end
x = single_seed_pagerank(P,0.85,1)
utils.graphplot(A,xy)
scatter(xy[:,1],xy[:,2],400*x+5,400*x,edgecolor="None",cmap="inferno")
#axis("off")
#colorbar()


    

    













    
Out[35]:





PyObject <matplotlib.collections.PathCollection object at 0x3243ea710>

In [36]:

    

using Interact


    

In [37]:

    

n = size(P,1)
f = figure()
@manipulate for alpha=0:0.01:0.99, seed=1:n; withfig(f) do
        x = single_seed_pagerank(P,alpha,seed)
        utils.graphplot(A,xy)
        scatter(xy[:,1],xy[:,2],25*max(log(n*x),0.25),max(log(n*x),0),edgecolor="None",cmap="inferno")
        axis("off")
        colorbar()
        #clim([0,sqrt(n)])
    end   
end


    

    












    












    
Out[37]:

In [38]:

    

close("all")


    

In [45]:

    

include("diffusions.jl")
include("utils.jl")
using DiffusionAlgorithms
using DiffusionTutorialUtilities

A,xy = utils.load_graph_and_coords("newman-netsci");
P = utils.normout!(A)
n = size(P,1)
# chebyshev point set of alphas
N=15; alphas = [0.999,0.995,0.99,0.95,0.9,0.85,0.8,0.75,0.66,0.5,0.25,0.1]
@manipulate for alpha=alphas, seed=1:n
    @time x = DiffusionAlgorithms.single_seed_pagerank_power(P',alpha,seed)
    #utils.cgraphplot(A,xy,0.33*max(log(n*x),0.75),max(log(n*x),0),utils._inferno_data)
    utils.cgraphplot(A,xy,0.1*log(1/(max(minimum(x),1.e-12))*x),log(x+1.e-12),
        utils._inferno_data)

end


    

    












    












    




WARNING: replacing module DiffusionAlgorithms
WARNING: replacing module DiffusionTutorialUtilities






    




  





    
Out[45]:













    




0.143544 seconds (39.54 k allocations: 1.836 MB)

In [46]:

    

include("diffusions.jl")
include("utils.jl")
using DiffusionAlgorithms
using DiffusionTutorialUtilities

A,xy = utils.load_graph_and_coords("U3A");
P = utils.normout!(A)
n = size(P,1)
# chebyshev point set of alphas
N=15; alphas = [0.999,0.995,0.99,0.95,0.9,0.85,0.8,0.75,0.66,0.5,0.25,0.1]
@manipulate for alpha=alphas, seed=1:n
    @time x = DiffusionAlgorithms.single_seed_pagerank_power(P',alpha,seed)
    #utils.cgraphplot(A,xy,0.33*max(log(n*x),0.75),max(log(n*x),0),utils._inferno_data)
    utils.cgraphplot(A,xy,0.1*log(1/(max(minimum(x),1.e-12))*x),log(x+1.e-12),
        utils._inferno_data)
end


    

    












    












    




WARNING: replacing module DiffusionAlgorithms
WARNING: replacing module DiffusionTutorialUtilities






    




  





    
Out[46]: