# These are the files to work with

with open('Output/synthetic_biology/synthetic_biology_JSON.txt', 'r') as f:      
    synth_wdfreq = json.load(f)
f.close()

with open('Output/synthetic_biology/synthetic_biology_nation.txt', 'r') as g:      
    synth_nation = json.load(g)
g.close()

with open('Output/synthetic_biology/synthetic_biology_year.txt', 'r') as h:      
    synth_year = json.load(h)
h.close()

print(type(data))

# Make a dictionary of word counters

import json

with open('Output/synthetic_biology/synthetic_biology_JSON.txt') as fg:
    word_counters = json.load(fg)
fg.close()

print("There are "+str(len(word_counters))+" indexed articles in this file.")

# Make a dictionary of word counters

import json

with open('Output/synthetic_biology/synthetic_biology_nation.txt') as fg:
    word_counters = json.load(fg)
fg.close()

print("There are "+str(len(word_counters))+" indexed articles in this file.")

# Take a look at the nation file records

begin_line = -1
end_line = 500

with open('Data/synthetic_biology_nation.txt') as j:
    for n,line in enumerate(j):
        if ((n > begin_line) and (n < end_line)):
            print(line.rstrip())
j.close()

import json
from pprint import pprint

with open('Output/synthetic_biology/synthetic_biology_JSON.txt') as data_file:
    data = json.load(data_file)

article1 = data['24838884']
article2 = data['26142061']
article3 = data['26132395']

print(article1.keys())

article_id = "23825301"

my_content = word_counters[article_id]

print("The nation in this article is "+str(my_content)+".")
print()

for n,key in enumerate(data):
    if (n < 3):
        print(key)

with open('Output/synthetic_biology/synthetic_biology_nation.txt') as data_file:
    data = json.load(data_file)

article1 = data['24838884']
article2 = data['26142061']
article3 = data['26132395']

article_list = [article1, article2, article3]
for article in article_list:
    print("The country is", article)

with open('Output/synthetic_biology/synthetic_biology_nation.txt') as data_file:
    data = json.load(data_file)

article = data['24838884']
print(article)
print()

with open('Output/synthetic_biology/synthetic_biology_JSON.txt') as fg:
    synth_supp = json.load(fg)
print(article_id)

# create a new network N
import networkx as nx

import itertools as it

N = nx.Graph()

for key in synth_nation:
    nation_list = synth_nation[key]
    print(key, nation_list)
    N.add_nodes_from(nation_list)
    l = list(it.combinations(nation_list,2))
    if (len(l) > 0):
        for e in l:
            a = e[0]
            b = e[1]
            if (N.has_edge(a,b)):
                N[a][b]['weight'] += 1
            else:
                N.add_edge(a,b,weight=1)

print(N.nodes())
print(N.edges())
N.remove_node('NONE')
N.remove_node('OTHER')

# Do a quick functional plot of the data 
import matplotlib.pyplot as plt

# set the graphic inline in screen
# set the figure size using plt dot figure
%matplotlib inline
plt.figure(figsize=(11,11))

# do some layouts
pos1=nx.fruchterman_reingold_layout(N)

# set the labels, draw the graph
nx.draw_networkx_labels(N,pos1)
nx.draw(N,pos1)

# Take a closer look at the frequency of collaboration between institutions

import matplotlib.pyplot as plt

# set the graphic inline in screen
# set the figure size
%matplotlib inline
plt.figure(figsize=(13,13))

# do some layouts
pos1=nx.fruchterman_reingold_layout(N)
pos2=nx.circular_layout(N)
pos3=nx.spring_layout(N)
pos4=nx.spectral_layout(N)

# set the labels, draw the graph
plt.subplot(2,2,1)
plt.title("Fruchterman Reingold Layout",fontsize=14)
nx.draw_networkx_labels(N,pos1)
nx.draw(N,pos1)

# set the labels, draw the graph
plt.subplot(2,2,2)
plt.title("Circular Layout",fontsize=14)
nx.draw_networkx_labels(N,pos2)
nx.draw(N,pos2)

# set the labels, draw the graph
plt.subplot(2,2,3)
plt.title("Spring Layout",fontsize=14)
nx.draw_networkx_labels(N,pos3)
nx.draw(N,pos3)

# set the labels, draw the graph
plt.subplot(2,2,4)
plt.title("Spectral Layout",fontsize=14)
nx.draw_networkx_labels(N,pos4)
nx.draw(N,pos4)

# Calculate Betweenness and Centrality in this network

cc=nx.eigenvector_centrality(N)
bc=nx.betweenness_centrality(N)
nx.set_node_attributes(N, 'betweenness', bc)
nx.set_node_attributes(N, 'centrality', cc)

# Determine the most centrally located nodes

import operator

sorted_cc = sorted(cc.items(), key=operator.itemgetter(1))
print("Below are the most centrally located nodes, ranked from least to greatest.")
print()
print(sorted_cc)

# Determine the most between nodes

import operator

sorted_bc = sorted(bc.items(), key=operator.itemgetter(1))
print("Below are the most between nodes, ranked from least to greatest")
print()
print(sorted_bc)

NN = N.copy()

nation_list_data = NN.nodes(data=True)
print(nation_list_data)

# Dereference the data and print in a friendly list
for nation,data in nation_list_data:
    print(nation)
    print('centrality:',data['centrality'])
    print('betweenness:',data['betweenness'])
    print()

# Save the eigenvector to a dictionary
cc1 = {}
for nation,data in nation_list_data:
    centrality = data['centrality']
    cc1[nation]=centrality

# Add some color
plt.figure(figsize=(25,15))
pos=nx.fruchterman_reingold_layout(NN)
nx.draw_networkx(NN,pos,nodelist=cc1.keys(),node_color='#FFFF00',alpha=0.30,font_size=8)

# This is a funky command that transforms the centrality measure to the node size and makes us ready to make bubble proportionate
ns = []

for c in cc1.values():
    # Scale this up by 10000
    ns.append(c*10000)
#print(cc1.keys())

# Change the size of the nodes

plt.figure(figsize=(30,30))
pos=nx.fruchterman_reingold_layout(NN)
d = nx.eigenvector_centrality(NN)
nx.draw_networkx(NN,nodelist=cc1.keys(),node_size=ns,node_color='#4169E1',alpha=0.20,font_size=12)

# Add the list of weights for each of the edges or connecting lines
weight_list = []
for u,v in NN.edges():
    w = NN[u][v]['weight']
    weight_list.append(w)

# Redo the graph with the weight of the line indicating the strength of the collaboration

plt.figure(figsize=(30,15))
pos=nx.fruchterman_reingold_layout(NN)
edges = NN.edges()
nx.draw_networkx(NN,pos,nodelist=cc1.keys(),node_size=ns,node_color='#FFFF00',edge_color='#e2b7a0',
                 alpha=0.35,font_size=10,width=weight_list)

# And make it round 

plt.figure(figsize=(20,20))
pos=nx.circular_layout(NN)
edges = NN.edges()
nx.draw_networkx(NN,pos2,nodelist=cc1.keys(),node_size=ns,node_color='#FF0000',edge_color='#e2b7a0',
                 alpha=0.35,font_size=10,width=weight_list)