How to make Choropleth map about Covid-19 Cases in New Zealand

Import the packages

from lxml.html import parse
from urllib.request import urlopen
from pandas.io.parsers import TextParser
import pandas as pd
from pandas import DataFrame, Series
import numpy as np
import folium

Get the cases data for mapping

Define the function to parse the table data from website

def parse_options_data(table):
    rows = table.findall('.//tr') # read out all the content in the table
    header = _unpack(rows[0], kind='th') # read out the header of the table
    data = [_unpack(r) for r in rows[1:]]  # read out the table
    return TextParser(data, names=header).get_chunk()

#reterive the Confirmed Covid-19 cases data from Health of Ministry website
url = 'https://www.health.govt.nz/our-work/diseases-and-conditions/covid-19-novel-coronavirus/covid-19-current-cases'
parsed = parse(urlopen(url))
doc = parsed.getroot()
tables = doc.findall('.//table') # to select all the tables on the target website
calls = tables[1] # select the target table
call_data = parse_options_data(calls)
df = DataFrame(call_data)# put the data into pandas dataframe

df.head()

	DHB	Number of cases	Change in last 24 hours
0	Auckland	180	4
1	Bay of Plenty	41	0
2	Canterbury	139	3
3	Capital and Coast	88	0
4	Counties Manukau	103	2

Treat the special character

df.DHB = list(i.replace('ā','a') for i in df.DHB)

df.tail()

	DHB	Number of cases	Change in last 24 hours
16	Wairarapa	8	0
17	Waitemata	200	5
18	West Coast	5	0
19	Whanganui	7	0
20	Total	1,366	17

Now we get the table of addresses and the cases No. data for mapping.

Then we need to get the location data so the program know where to draw.

Geolocation data

read out the center of each DHB cities

df3 = pd.read_csv('DHBLocation.csv')
df3

	DHB	Y	X
0	Northland	173.825907	-35.489931
1	Waitemata	174.546651	-36.563964
2	Auckland	175.156221	-36.517984
3	Counties Manukau	174.932391	-37.187695
4	Waikato	175.356931	-38.048669
5	Lakes	176.123138	-38.653552
6	Bay of Plenty	176.918989	-38.091072
7	Tairawhiti	177.921101	-38.281284
8	Taranaki	174.467469	-39.315027
9	Hawke's Bay	176.796960	-39.412258
10	Whanganui	175.519157	-39.609703
11	MidCentral	175.810108	-40.382262
12	Hutt Valley	175.051906	-41.160282
13	Capital and Coast	174.897246	-41.092649
14	Wairarapa	175.628010	-41.099410
15	Nelson Marlborough	173.137270	-41.362035
16	West Coast	170.870664	-42.794808
17	Canterbury	172.366210	-43.075242
18	South Canterbury	170.656863	-44.146524
19	Southern	168.789955	-45.442032

Merge the table of case No, and location based on DHB cities

df4 = pd.merge(df3, df, how='left', on='DHB')

df4.head()

	DHB	Y	X	Number of cases	Change in last 24 hours
0	Northland	173.825907	-35.489931	26	1
1	Waitemata	174.546651	-36.563964	200	5
2	Auckland	175.156221	-36.517984	180	4
3	Counties Manukau	174.932391	-37.187695	103	2
4	Waikato	175.356931	-38.048669	177	0

df4['Number of cases'] = df4['Number of cases'].astype(int) # change the values to int

Mapping

create a list of linear spacing for color range

nzta_geo = 'DHB.geojson' #this file is taken from the LINZ website dated on 2016, it including the polygon information of each DHB cities

# create a numpy array of length 6 and has linear spacing from the minium cases No. to the maximum cases No.
import math
threshold_scale = np.linspace(df4['Number of cases'].min(),
                              df4['Number of cases'].max(),
                              6, dtype=int)
threshold_scale = threshold_scale.tolist() # change the numpy array to a list


threshold_scale[-1] = threshold_scale[-1] + 1 # make sure that the last value of the list is greater than the maximum cases No.
threshold_scale

[3, 44, 85, 127, 168, 211]

Choropleth mapping

#Choropleth mapping
#creat the map object
world_map = folium.Map(location=[-39.58,175.3698], zoom_start=5.5, tiles='Stamen Terrain')
folium.Choropleth(
    geo_data = nzta_geo,
    data=df4, #data source
    columns=['DHB', 'Number of cases'],  # data columns
    key_on='feature.properties.DHB2015_Na', # the geometry infomation on the geojson data
    threshold_scale=threshold_scale,
    fill_color='YlOrRd', #color
    fill_opacity=0.7,
    line_opacity=0.2,
    legend_name='Confirmed Coronavirus Cases in NZ(Click icon check more details)',
#     reset=True
).add_to(world_map)

folium.LayerControl().add_to(world_map)

<folium.map.LayerControl at 0x1cebe5c6438>

Add the case No. icon on the map

#icon mapping
for i in range(len(df4.DHB)):

    folium.Marker(
        location=df4.iloc[i,[2,1]],
        popup=(r'{0};\nTTL: {1};\n 24H: {2}'.format(df4.iloc[i,0],df4.iloc[i,3],df4.iloc[i,4])),
        icon=folium.Icon(color='green')
    ).add_to(world_map)

Plotting the map out

# display map
world_map

Save the map to local disk as a website, and we can share with friend or post online.

import datetime
today_date=str(datetime.date.today())
world_map.save('Covid-19 Cases in New Zealand-map{}.html'.format(today_date))

final map result:https://fordy7014.github.io/YangWeb/

Reference

https://www.stats.govt.nz/

https://www.health.govt.nz/our-work/diseases-and-conditions/covid-19-novel-coronavirus/covid-19-current-situation/covid-19-current-cases