In [1]:

    

import pandas as pd, numpy as np
import matplotlib.pyplot as plt
%matplotlib inline
from jupyterthemes import jtplot
jtplot.style('solarized-light')


    

In [2]:

    

w=pd.read_html('https://en.wikipedia.org/wiki/List_of_the_most_intense_tropical_cyclones')


    

In [3]:

    

n_atl=w[0].T.set_index(0).T.dropna()


    

In [4]:

    

e_pac=w[1].T.set_index(0).T.dropna()


    

In [5]:

    

nw_pac=w[2].T.set_index(0).T.dropna()


    

In [6]:

    

n_ind=w[3].T.set_index(0).T.dropna()


    

In [7]:

    

sw_ind=w[4].T.set_index(0).T.dropna()


    

In [8]:

    

aus=w[5].T.set_index(0).T.dropna()


    

In [9]:

    

s_pac=w[6].T.set_index(0).T.dropna()


    

In [10]:

    

s_atl=w[7].T.set_index(0).T.dropna()


    

In [11]:

    

df=[n_atl,s_atl,e_pac,nw_pac,n_ind,sw_ind,aus,s_pac]


    

In [27]:

    

data=[]
regions=['North Atlantic','South Atlantic','East Pacific','Northwest Pacific','North Indian','South Indian','Australian','South Pacific']
for j in range(7):
    for i in df[j].T.iteritems():
        d={"region":regions[j]}
        d["name"]=i[1][0].replace('"','')
        d["pressure"]=int(i[1][3][i[1][3].find('hPa')-4:i[1][3].find('hPa')].strip())
        if u'\u2013' in i[1][1]:
            d["year"]=int(i[1][1][:i[1][1].find(u'\u2013')])
        elif u'-' in i[1][1]:
            d["year"]=int(i[1][1][:i[1][1].find(u'-')])
        else: d["year"]=int(i[1][1])
        k=i[1][2][i[1][2].find('!')+1:i[1][2][i[1][2].find('!')+1:].strip().find(' ')].strip()
        if u'\u2660' in k:
            d["speed"]=int(k[k.find(u'\u2660')+1:k.find(u'\xa0')])
        elif k!='':d["speed"]=int(k)
        data.append(d)


    

In [29]:

    

import json


    

In [30]:

    

file('data.json','w').write(json.dumps(data))