X-Git-Url: https://git.njae.me.uk/?a=blobdiff_plain;f=deaths_import.md;fp=deaths_import.md;h=f704518478859d3954c45c7dca1093fa3b92fa60;hb=5afedd66506be7575034ae6deebcfaa7c2ced978;hp=0000000000000000000000000000000000000000;hpb=4abff18d7988bdea04a267a08a0792ba570fe0bd;p=covid19.git

diff --git a/deaths_import.md b/deaths_import.md
new file mode 100644
index 0000000..f704518
--- /dev/null
+++ b/deaths_import.md
@@ -0,0 +1,1017 @@
+---
+jupyter:
+  jupytext:
+    formats: ipynb,md
+    text_representation:
+      extension: .md
+      format_name: markdown
+      format_version: '1.2'
+      jupytext_version: 1.9.1
+  kernelspec:
+    display_name: Python 3
+    language: python
+    name: python3
+---
+
+<!-- #region Collapsed="false" -->
+Data from:
+
+* [Office of National Statistics](https://www.ons.gov.uk/peoplepopulationandcommunity/birthsdeathsandmarriages/deaths/datasets/weeklyprovisionalfiguresondeathsregisteredinenglandandwales) (Endland and Wales) Weeks start on a Saturday.
+* [Northern Ireland Statistics and Research Agency](https://www.nisra.gov.uk/publications/weekly-deaths) (Northern Ireland). Weeks start on a Saturday. Note that the week numbers don't match the England and Wales data.
+* [National Records of Scotland](https://www.nrscotland.gov.uk/statistics-and-data/statistics/statistics-by-theme/vital-events/general-publications/weekly-and-monthly-data-on-births-and-deaths/weekly-data-on-births-and-deaths) (Scotland). Note that Scotland uses ISO8601 week numbers, which start on a Monday.
+
+<!-- #endregion -->
+
+```python Collapsed="false"
+import itertools
+import collections
+import json
+import pandas as pd
+import numpy as np
+from scipy.stats import gmean
+import datetime
+
+import matplotlib as mpl
+import matplotlib.pyplot as plt
+%matplotlib inline
+
+from sqlalchemy.types import Integer, Text, String, DateTime, Float
+from sqlalchemy import create_engine
+%load_ext sql
+```
+
+```python Collapsed="false"
+connection_string = 'postgresql://covid:3NbjJTkT63@localhost/covid'
+```
+
+```python Collapsed="false"
+%sql $connection_string
+```
+
+```python Collapsed="false"
+conn = create_engine(connection_string)
+```
+
+```python Collapsed="false"
+england_wales_filename = 'uk-deaths-data/publishedweek532020.xlsx'
+```
+
+```sql Collapsed="false"
+drop table if exists all_causes_deaths;
+create table all_causes_deaths (
+    week integer,
+    year integer,
+    date_up_to date,
+    nation varchar(20),
+    deaths integer,
+    CONSTRAINT week_nation PRIMARY KEY(year, week, nation)
+);
+```
+
+```python Collapsed="false"
+raw_data_2015 = pd.read_csv('uk-deaths-data/Weekly_Deaths_NI_2015.csv', 
+                       parse_dates=[1, 2], dayfirst=True,
+#                       index_col=0,
+                      header=[0, 1]
+                           )
+dh15i = raw_data_2015.iloc[:, [0, 3]]
+dh15i.set_index(dh15i.columns[0], inplace=True)
+dh15i.columns = ['total_2015']
+dh15i.tail()
+```
+
+```python Collapsed="false"
+raw_data_2015.head()
+```
+
+```python Collapsed="false"
+rd = raw_data_2015.iloc[:, [0, 2, 3]].droplevel(1, axis=1).rename(
+    columns={'Week Ends (Friday)': 'date_up_to', 'Total Number of Deaths Registered in Week (2015P)': 'deaths',
+            'Registration Week': 'week'}
+    )
+rd['year'] = 2015
+rd['nation'] = 'Northern Ireland'
+rd.head()
+```
+
+```python Collapsed="false"
+rd.to_sql(
+    'all_causes_deaths',
+    conn,
+    if_exists='append',
+    index=False)
+```
+
+```python Collapsed="false"
+%sql select * from all_causes_deaths limit 10
+```
+
+```python Collapsed="false"
+raw_data_2016 = pd.read_csv('uk-deaths-data/Weekly_Deaths_NI_2016.csv', 
+                        parse_dates=[1, 2], dayfirst=True,
+#                       index_col=0,
+                      header=[0, 1]
+                           )
+raw_data_2016.head()
+# dh16i = raw_data_2016.iloc[:, [2]]
+# dh16i.columns = ['total_2016']
+# # dh16i.head()
+dh16i = raw_data_2016.iloc[:, [0, 3]]
+dh16i.set_index(dh16i.columns[0], inplace=True)
+dh16i.columns = ['total_2016']
+dh16i.tail()
+```
+
+```python Collapsed="false"
+rd = raw_data_2016.iloc[:, [0, 2, 3]].droplevel(1, axis=1).rename(
+    columns={'Week Ends (Friday)': 'date_up_to', 'Total Number of Deaths Registered in Week (2016P)': 'deaths',
+            'Registration Week': 'week'}
+    )
+rd['year'] = 2016
+rd['nation'] = 'Northern Ireland'
+rd.head()
+```
+
+```python Collapsed="false"
+rd.to_sql(
+    'all_causes_deaths',
+    conn,
+    if_exists='append',
+    index=False)
+```
+
+```python Collapsed="false"
+%sql select year, nation, count(date_up_to) from all_causes_deaths group by (year, nation)
+```
+
+```python Collapsed="false"
+raw_data_2017 = pd.read_csv('uk-deaths-data/Weekly_Deaths_NI_2017.csv', 
+                        parse_dates=[1, 2], dayfirst=True,
+#                       index_col=0,
+                      header=[0, 1]
+                           )
+raw_data_2017.head()
+dh17i = raw_data_2017.iloc[:, [0, 3]]
+dh17i.set_index(dh17i.columns[0], inplace=True)
+dh17i.columns = ['total_2017']
+dh17i.tail()
+```
+
+```python Collapsed="false"
+rd = raw_data_2017.iloc[:, [0, 2, 3]].droplevel(1, axis=1).rename(
+    columns={'Week Ends (Friday)': 'date_up_to', 'Total Number of Deaths Registered in Week (2017P)': 'deaths',
+            'Registration Week': 'week'}
+    )
+rd['year'] = 2017
+rd['nation'] = 'Northern Ireland'
+rd.head()
+```
+
+```python Collapsed="false"
+rd.to_sql(
+    'all_causes_deaths',
+    conn,
+    if_exists='append',
+    index=False)
+```
+
+```python Collapsed="false"
+%sql select year, nation, count(date_up_to) from all_causes_deaths group by (year, nation)
+```
+
+```python Collapsed="false"
+raw_data_2018 = pd.read_csv('uk-deaths-data/Weekly_Deaths_NI_2018.csv', 
+                        parse_dates=[1, 2], dayfirst=True,
+#                       index_col=0,
+                      header=[0, 1]
+                           )
+raw_data_2018.head()
+dh18i = raw_data_2018.iloc[:, [0, 3]]
+dh18i.set_index(dh18i.columns[0], inplace=True)
+dh18i.columns = ['total_2018']
+dh18i.tail()
+```
+
+```python Collapsed="false"
+rd = raw_data_2018.iloc[:, [0, 2, 3]].droplevel(1, axis=1).rename(
+    columns={'Week Ends (Friday)': 'date_up_to', 'Total Number of Deaths Registered in Week (2018P)': 'deaths',
+            'Registration Week': 'week'}
+    )
+rd['year'] = 2018
+rd['nation'] = 'Northern Ireland'
+rd.head()
+```
+
+```python Collapsed="false"
+rd.to_sql(
+    'all_causes_deaths',
+    conn,
+    if_exists='append',
+    index=False)
+```
+
+```python Collapsed="false"
+%sql select year, nation, count(date_up_to) from all_causes_deaths group by (year, nation)
+```
+
+```python Collapsed="false"
+raw_data_2019 = pd.read_csv('uk-deaths-data/Weekly_Deaths_NI_2019.csv', 
+                        parse_dates=[1, 2], dayfirst=True,
+#                       index_col=0,
+                      header=[0, 1]
+                           )
+raw_data_2019.head()
+dh19i = raw_data_2019.iloc[:, [0, 3]]
+dh19i.set_index(dh19i.columns[0], inplace=True)
+dh19i.columns = ['total_2019']
+dh19i.tail()
+```
+
+```python Collapsed="false"
+rd = raw_data_2019.iloc[:, [0, 2, 3]].droplevel(1, axis=1).rename(
+    columns={'Week Ends (Friday)': 'date_up_to', 'Total Number of Deaths Registered in Week (2019P)': 'deaths',
+            'Registration Week': 'week'}
+    )
+rd['year'] = 2019
+rd['nation'] = 'Northern Ireland'
+rd.head()
+```
+
+```python Collapsed="false"
+rd.to_sql(
+    'all_causes_deaths',
+    conn,
+    if_exists='append',
+    index=False)
+```
+
+```python Collapsed="false"
+%sql select year, nation, count(date_up_to) from all_causes_deaths group by (year, nation)
+```
+
+```python Collapsed="false"
+raw_data_2020_i = pd.read_csv('uk-deaths-data/Weekly_Deaths_NI_2020.csv', 
+                        parse_dates=[1], dayfirst=True,
+                      header=[0, 1]
+                           )
+raw_data_2020_i.head()
+```
+
+```python Collapsed="false"
+rd = raw_data_2020_i.iloc[:, [0, 1, 2]].droplevel(1, axis=1).rename(
+    columns={'Week Ending (Friday)': 'date_up_to', 'Total Number of Deaths Registered in Week (2020P)': 'deaths',
+            'Registration Week': 'week'}
+    )
+rd['year'] = 2020
+rd['nation'] = 'Northern Ireland'
+rd.head()
+```
+
+```python Collapsed="false"
+rd.tail()
+```
+
+```python Collapsed="false"
+rd.to_sql(
+    'all_causes_deaths',
+    conn,
+    if_exists='append',
+    index=False)
+```
+
+```python Collapsed="false"
+%sql select year, nation, count(date_up_to) from all_causes_deaths group by (year, nation) order by nation, year
+```
+
+```python
+raw_data_2020_i.set_index(raw_data_2020_i.columns[0], inplace=True)
+raw_data_2020_i.tail()
+```
+
+```python Collapsed="false"
+datetime.datetime.now().isocalendar()
+```
+
+```python Collapsed="false"
+datetime.datetime.fromisocalendar(2021, 3, 1)
+```
+
+```python Collapsed="false"
+
+```
+
+```python Collapsed="false"
+raw_data_s = pd.read_csv('uk-deaths-data/weekly-deaths-scotland.csv', 
+                      index_col=0,
+                      header=0,
+                        skiprows=2
+                           )
+# raw_data_s
+```
+
+```python Collapsed="false"
+deaths_headlines_s = raw_data_s[reversed('2015 2016 2017 2018 2019 2020'.split())]
+deaths_headlines_s.columns = ['total_' + c for c in deaths_headlines_s.columns]
+deaths_headlines_s.reset_index(drop=True, inplace=True)
+deaths_headlines_s.index = deaths_headlines_s.index + 1
+deaths_headlines_s
+```
+
+```python
+%sql select * from all_causes_deaths limit 5
+```
+
+```python Collapsed="false"
+for year, ser in deaths_headlines_s.items():
+    year_i = int(year[-4:])
+#     print(year_i)
+    for week, deaths in ser.dropna().iteritems():
+#         print(datetime.date.fromisocalendar(year_i, week, 7), deaths)
+        dut = datetime.date.fromisocalendar(year_i, week, 7)
+        %sql insert into all_causes_deaths(week, year, date_up_to, nation, deaths) values ({week}, {year_i}, :dut, 'Scotland', {deaths})
+```
+
+```python
+%sql select year, nation, count(date_up_to) from all_causes_deaths group by (year, nation) order by nation, year
+```
+
+```python
+%sql select year, nation, date_up_to from all_causes_deaths where week=3 order by year, nation
+```
+
+```python Collapsed="false"
+eng_xls = pd.read_excel(england_wales_filename, 
+                        sheet_name="Weekly figures 2020",
+                        skiprows=[0, 1, 2, 3],
+                        header=0,
+                        index_col=[1]
+                       ).iloc[:91].T
+eng_xls
+```
+
+```python Collapsed="false"
+# eng_xls_columns
+```
+
+```python Collapsed="false"
+eng_xls_columns = list(eng_xls.columns)
+
+for i, c in enumerate(eng_xls_columns):
+#     print(i, c, type(c), isinstance(c, float))
+    if isinstance(c, float) and np.isnan(c):
+        if eng_xls.iloc[0].iloc[i] is not pd.NaT:
+            eng_xls_columns[i] = eng_xls.iloc[0].iloc[i]
+
+# np.isnan(eng_xls_columns[0])
+# eng_xls_columns
+
+eng_xls.columns = eng_xls_columns
+# eng_xls.columns
+```
+
+```python
+eng_xls
+```
+
+```python
+rd = eng_xls.iloc[1:][['Week ended', 'Wales']].reset_index(level=0).rename(
+    columns={'Week ended': 'date_up_to', 'Wales': 'deaths',
+            'index': 'week'}
+    )
+rd['year'] = 2020
+rd['nation'] = 'Wales'
+rd.head()
+```
+
+```python
+rd.to_sql(
+    'all_causes_deaths',
+    conn,
+    if_exists='append',
+    index=False)
+```
+
+```python Collapsed="false"
+eng_xls = eng_xls.iloc[1:]
+eng_xls['England deaths'] = eng_xls.loc[:, 'Total deaths, all ages'] - eng_xls.loc[:, 'Wales']
+```
+
+```python
+eng_xls.head()
+```
+
+```python
+rd = eng_xls[['Week ended', 'England deaths']].reset_index(level=0).rename(
+    columns={'Week ended': 'date_up_to', 'England deaths': 'deaths',
+            'index': 'week'}
+    )
+rd['year'] = 2020
+rd['nation'] = 'England'
+rd.head()
+```
+
+```python
+%sql delete from all_causes_deaths where nation = 'England'
+```
+
+```python
+rd.to_sql(
+    'all_causes_deaths',
+    conn,
+    if_exists='append',
+    index=False)
+```
+
+```python
+rd.tail()
+```
+
+```python
+%sql select year, nation, count(date_up_to) from all_causes_deaths group by (year, nation) order by nation, year
+```
+
+```python Collapsed="false"
+# raw_data_2020 = pd.read_csv('uk-deaths-data/publishedweek272020.csv', 
+#                        parse_dates=[1], dayfirst=True,
+#                       index_col=0,
+#                       header=[0, 1])
+```
+
+```python Collapsed="false"
+
+```
+
+```python Collapsed="false"
+# raw_data_2020.head()
+```
+
+```python Collapsed="false"
+# raw_data_2020['W92000004', 'Wales']
+```
+
+```python Collapsed="false"
+raw_data_2019 = pd.read_csv('uk-deaths-data/publishedweek522019.csv', 
+                       parse_dates=[1], dayfirst=True,
+#                       index_col=0,
+                      header=[0, 1])
+# raw_data_2019.head()
+```
+
+```python
+rdew = raw_data_2019.iloc[:, [0, 1, 2, -1]].droplevel(axis=1, level=1)
+rdew.head()
+```
+
+```python
+rd = rdew.drop(columns=['Total deaths, all ages']).rename(
+    columns={'Week ended': 'date_up_to', 'W92000004': 'deaths',
+            'Week number': 'week'}
+    )
+rd['year'] = 2019
+rd['nation'] = 'Wales'
+rd.head()
+```
+
+```python
+rd.to_sql(
+    'all_causes_deaths',
+    conn,
+    if_exists='append',
+    index=False)
+```
+
+```python
+rd = rdew.loc[:, ['Week ended','Week number']]
+rd['deaths'] = rdew['Total deaths, all ages'] - rdew['W92000004']
+rd = rd.rename(
+    columns={'Week ended': 'date_up_to', 'Week number': 'week'}
+    )
+rd['year'] = 2019
+rd['nation'] = 'England'
+rd.head()
+```
+
+```python
+rd.to_sql(
+    'all_causes_deaths',
+    conn,
+    if_exists='append',
+    index=False)
+```
+
+```python
+%sql select year, nation, count(date_up_to) from all_causes_deaths group by (year, nation) order by nation, year
+```
+
+```python Collapsed="false"
+raw_data_2018 = pd.read_csv('uk-deaths-data/publishedweek522018.csv', 
+                       parse_dates=[1], dayfirst=True,
+#                       index_col=0,
+                      header=[0, 1])
+# raw_data_2018.head()
+```
+
+```python
+rdew = raw_data_2018.iloc[:, [0, 1, 2, -1]].droplevel(axis=1, level=1)
+rdew.head()
+```
+
+```python
+rd = rdew.drop(columns=['Total deaths, all ages']).rename(
+    columns={'Week ended': 'date_up_to', 'W92000004': 'deaths',
+            'Week number': 'week'}
+    )
+rd['year'] = 2018
+rd['nation'] = 'Wales'
+rd.head()
+```
+
+```python
+rd.to_sql(
+    'all_causes_deaths',
+    conn,
+    if_exists='append',
+    index=False)
+```
+
+```python
+rd = rdew.loc[:, ['Week ended','Week number']]
+rd['deaths'] = rdew['Total deaths, all ages'] - rdew['W92000004']
+rd = rd.rename(
+    columns={'Week ended': 'date_up_to', 'Week number': 'week'}
+    )
+rd['year'] = 2018
+rd['nation'] = 'England'
+rd.head()
+```
+
+```python
+rd.to_sql(
+    'all_causes_deaths',
+    conn,
+    if_exists='append',
+    index=False)
+```
+
+```python
+%sql select year, nation, count(date_up_to) from all_causes_deaths group by (year, nation) order by nation, year
+```
+
+```python Collapsed="false"
+raw_data_2017 = pd.read_csv('uk-deaths-data/publishedweek522017.csv', 
+                       parse_dates=[1], dayfirst=True,
+#                       index_col=0,
+                      header=[0, 1])
+# raw_data_2017.head()
+```
+
+```python
+rdew = raw_data_2017.iloc[:, [0, 1, 2, -1]].droplevel(axis=1, level=1)
+rdew.head()
+```
+
+```python
+rd = rdew.drop(columns=['Total deaths, all ages']).rename(
+    columns={'Week ended': 'date_up_to', 'W92000004': 'deaths',
+            'Week number': 'week'}
+    )
+rd['year'] = 2017
+rd['nation'] = 'Wales'
+rd.head()
+```
+
+```python
+rd.to_sql(
+    'all_causes_deaths',
+    conn,
+    if_exists='append',
+    index=False)
+```
+
+```python
+rd = rdew.loc[:, ['Week ended','Week number']]
+rd['deaths'] = rdew['Total deaths, all ages'] - rdew['W92000004']
+rd = rd.rename(
+    columns={'Week ended': 'date_up_to', 'Week number': 'week'}
+    )
+rd['year'] = 2017
+rd['nation'] = 'England'
+rd.head()
+```
+
+```python
+rd.to_sql(
+    'all_causes_deaths',
+    conn,
+    if_exists='append',
+    index=False)
+```
+
+```python
+%sql select year, nation, count(date_up_to) from all_causes_deaths group by (year, nation) order by nation, year
+```
+
+```python
+
+```
+
+```python Collapsed="false"
+raw_data_2016 = pd.read_csv('uk-deaths-data/publishedweek522016.csv', 
+                       parse_dates=[1], dayfirst=True,
+#                       index_col=0,
+                      header=[0, 1])
+# raw_data_2016.head()
+```
+
+```python
+raw_data_2016.head()
+```
+
+```python
+rdew = raw_data_2016.iloc[:, [0, 1, 2, -1]].droplevel(axis=1, level=1)
+rdew.head()
+```
+
+```python
+rd = rdew.drop(columns=['Total deaths, all ages']).rename(
+    columns={'Week ended': 'date_up_to', 'W92000004': 'deaths',
+            'Week number': 'week'}
+    )
+rd['year'] = 2016
+rd['nation'] = 'Wales'
+rd.head()
+```
+
+```python
+rd.to_sql(
+    'all_causes_deaths',
+    conn,
+    if_exists='append',
+    index=False)
+```
+
+```python
+rd = rdew.loc[:, ['Week ended','Week number']]
+rd['deaths'] = rdew['Total deaths, all ages'] - rdew['W92000004']
+rd = rd.rename(
+    columns={'Week ended': 'date_up_to', 'Week number': 'week'}
+    )
+rd['year'] = 2016
+rd['nation'] = 'England'
+rd.head()
+```
+
+```python
+rd.to_sql(
+    'all_causes_deaths',
+    conn,
+    if_exists='append',
+    index=False)
+```
+
+```python
+ %sql select year, nation, count(date_up_to) from all_causes_deaths group by (year, nation) order by nation, year
+```
+
+```python Collapsed="false"
+raw_data_2015 = pd.read_csv('uk-deaths-data/publishedweek2015.csv', 
+                       parse_dates=[1], dayfirst=True,
+#                       index_col=0,
+                      header=[0, 1])
+# raw_data_2015.head()
+```
+
+```python
+rdew = raw_data_2015.iloc[:, [0, 1, 2, -1]].droplevel(axis=1, level=1)
+rdew.head()
+```
+
+```python
+rd = rdew.drop(columns=['Total deaths, all ages']).rename(
+    columns={'Week ended': 'date_up_to', 'W92000004': 'deaths',
+            'Week number': 'week'}
+    )
+rd['year'] = 2015
+rd['nation'] = 'Wales'
+rd.head()
+```
+
+```python
+rd.to_sql(
+    'all_causes_deaths',
+    conn,
+    if_exists='append',
+    index=False)
+```
+
+```python
+rd = rdew.loc[:, ['Week ended','Week number']]
+rd['deaths'] = rdew['Total deaths, all ages'] - rdew['W92000004']
+rd = rd.rename(
+    columns={'Week ended': 'date_up_to', 'Week number': 'week'}
+    )
+rd['year'] = 2015
+rd['nation'] = 'England'
+rd.head()
+```
+
+```python
+rd.to_sql(
+    'all_causes_deaths',
+    conn,
+    if_exists='append',
+    index=False)
+```
+
+```python
+%sql select year, nation, count(date_up_to) from all_causes_deaths group by (year, nation) order by year, nation
+```
+
+```sql magic_args="res << select week, year, deaths"
+from all_causes_deaths
+where nation = 'England'
+```
+
+```python
+deaths_headlines_e = res.DataFrame().pivot(index='week', columns='year', values='deaths')
+deaths_headlines_e
+```
+
+```python
+
+```
+
+```sql magic_args="res << select week, year, deaths"
+from all_causes_deaths
+where nation = 'Scotland'
+```
+
+```python
+deaths_headlines_s = res.DataFrame().pivot(index='week', columns='year', values='deaths')
+deaths_headlines_s
+```
+
+```sql magic_args="res << select week, year, deaths"
+from all_causes_deaths
+where nation = 'Wales'
+```
+
+```python
+deaths_headlines_w = res.DataFrame().pivot(index='week', columns='year', values='deaths')
+deaths_headlines_w
+```
+
+```sql magic_args="res << select week, year, deaths"
+from all_causes_deaths
+where nation = 'Northern Ireland'
+```
+
+```python
+deaths_headlines_i = res.DataFrame().pivot(index='week', columns='year', values='deaths')
+deaths_headlines_i
+```
+
+```python Collapsed="false"
+deaths_headlines = deaths_headlines_e + deaths_headlines_w + deaths_headlines_i + deaths_headlines_s
+deaths_headlines
+```
+
+```python
+deaths_headlines_e.columns
+```
+
+```python
+deaths_headlines_e['previous_mean'] = deaths_headlines_e[[int(y) for y in '2019 2018 2017 2016 2015'.split()]].apply(np.mean, axis=1)
+deaths_headlines_w['previous_mean'] = deaths_headlines_w[[int(y) for y in '2019 2018 2017 2016 2015'.split()]].apply(np.mean, axis=1)
+deaths_headlines_s['previous_mean'] = deaths_headlines_s[[int(y) for y in '2019 2018 2017 2016 2015'.split()]].apply(np.mean, axis=1)
+deaths_headlines_i['previous_mean'] = deaths_headlines_i[[int(y) for y in '2019 2018 2017 2016 2015'.split()]].apply(np.mean, axis=1)
+deaths_headlines['previous_mean'] = deaths_headlines[[int(y) for y in '2019 2018 2017 2016 2015'.split()]].apply(np.mean, axis=1)
+deaths_headlines
+```
+
+```python Collapsed="false"
+deaths_headlines[[2020, 2019, 2018, 2017, 2016, 2015]].plot(figsize=(14, 8))
+```
+
+```python Collapsed="false"
+deaths_headlines[[2020, 'previous_mean']].plot(figsize=(10, 8))
+```
+
+```python Collapsed="false"
+deaths_headlines_i.plot()
+```
+
+```python
+deaths_headlines[2020].sum() - deaths_headlines.previous_mean.sum()
+```
+
+```python Collapsed="false"
+# Radar plot code taken from example at https://stackoverflow.com/questions/42878485/getting-matplotlib-radar-plot-with-pandas#
+
+dhna = deaths_headlines.dropna()
+
+fig = plt.figure(figsize=(10, 10))
+ax = fig.add_subplot(111, projection="polar")
+
+theta = np.roll(
+    np.flip(
+        np.arange(len(dhna))/float(len(dhna))*2.*np.pi),
+    14)
+# l15, = ax.plot(theta, deaths_headlines['total_2015'], color="#b56363", label="2015") # 0
+# l16, = ax.plot(theta, deaths_headlines['total_2016'], color="#a4b563", label="2016") # 72
+# l17, = ax.plot(theta, deaths_headlines['total_2017'], color="#63b584", label="2017") # 144
+# l18, = ax.plot(theta, deaths_headlines['total_2018'], color="#6384b5", label="2018") # 216
+# l19, = ax.plot(theta, deaths_headlines['total_2019'], color="#a4635b", label="2019") # 288
+l15, = ax.plot(theta, dhna[2015], color="#e47d7d", label="2015") # 0
+l16, = ax.plot(theta, dhna[2016], color="#afc169", label="2016") # 72 , d0e47d
+l17, = ax.plot(theta, dhna[2017], color="#7de4a6", label="2017") # 144
+l18, = ax.plot(theta, dhna[2018], color="#7da6e4", label="2018") # 216
+l19, = ax.plot(theta, dhna[2019], color="#d07de4", label="2019") # 288
+
+lmean, = ax.plot(theta, dhna['previous_mean'], color="black", linestyle='dashed', label="mean")
+
+l20, = ax.plot(theta, dhna[2020], color="red", label="2020")
+
+# deaths_headlines.total_2019.plot(ax=ax)
+
+def _closeline(line):
+    x, y = line.get_data()
+    x = np.concatenate((x, [x[0]]))
+    y = np.concatenate((y, [y[0]]))
+    line.set_data(x, y)
+
+[_closeline(l) for l in [l19, l18, l17, l16, l15, lmean]]
+
+
+ax.set_xticks(theta)
+ax.set_xticklabels(dhna.index)
+plt.legend()
+plt.title("Deaths by week over years, all UK")
+plt.savefig('deaths-radar.png')
+plt.show()
+```
+
+<!-- #region Collapsed="false" -->
+# Plots for UK nations
+<!-- #endregion -->
+
+```python Collapsed="false"
+# Radar plot code taken from example at https://stackoverflow.com/questions/42878485/getting-matplotlib-radar-plot-with-pandas#
+
+fig = plt.figure(figsize=(10, 10))
+ax = fig.add_subplot(111, projection="polar")
+
+theta = np.roll(
+    np.flip(
+        np.arange(len(deaths_headlines_e))/float(len(deaths_headlines_e))*2.*np.pi),
+    14)
+l15, = ax.plot(theta, deaths_headlines_e[2015], color="#e47d7d", label="2015") # 0
+l16, = ax.plot(theta, deaths_headlines_e[2016], color="#afc169", label="2016") # 72 , d0e47d
+l17, = ax.plot(theta, deaths_headlines_e[2017], color="#7de4a6", label="2017") # 144
+l18, = ax.plot(theta, deaths_headlines_e[2018], color="#7da6e4", label="2018") # 216
+l19, = ax.plot(theta, deaths_headlines_e[2019], color="#d07de4", label="2019") # 288
+
+lmean, = ax.plot(theta, deaths_headlines_e['previous_mean'], color="black", linestyle='dashed', label="mean")
+
+l20, = ax.plot(theta, deaths_headlines_e[2020], color="red", label="2020")
+
+# deaths_headlines.total_2019.plot(ax=ax)
+
+def _closeline(line):
+    x, y = line.get_data()
+    x = np.concatenate((x, [x[0]]))
+    y = np.concatenate((y, [y[0]]))
+    line.set_data(x, y)
+
+[_closeline(l) for l in [l19, l18, l17, l16, l15, lmean]]
+
+
+ax.set_xticks(theta)
+ax.set_xticklabels(deaths_headlines_e.index)
+plt.legend()
+plt.title("Deaths by week over years, England")
+plt.savefig('deaths-radar_england.png')
+plt.show()
+```
+
+```python Collapsed="false"
+# Radar plot code taken from example at https://stackoverflow.com/questions/42878485/getting-matplotlib-radar-plot-with-pandas#
+
+fig = plt.figure(figsize=(10, 10))
+ax = fig.add_subplot(111, projection="polar")
+
+theta = np.roll(
+    np.flip(
+        np.arange(len(deaths_headlines_w))/float(len(deaths_headlines_w))*2.*np.pi),
+    14)
+l15, = ax.plot(theta, deaths_headlines_w[2015], color="#e47d7d", label="2015") # 0
+l16, = ax.plot(theta, deaths_headlines_w[2016], color="#afc169", label="2016") # 72 , d0e47d
+l17, = ax.plot(theta, deaths_headlines_w[2017], color="#7de4a6", label="2017") # 144
+l18, = ax.plot(theta, deaths_headlines_w[2018], color="#7da6e4", label="2018") # 216
+l19, = ax.plot(theta, deaths_headlines_w[2019], color="#d07de4", label="2019") # 288
+
+lmean, = ax.plot(theta, deaths_headlines_w['previous_mean'], color="black", linestyle='dashed', label="mean")
+
+l20, = ax.plot(theta, deaths_headlines_w[2020], color="red", label="2020")
+
+
+def _closeline(line):
+    x, y = line.get_data()
+    x = np.concatenate((x, [x[0]]))
+    y = np.concatenate((y, [y[0]]))
+    line.set_data(x, y)
+
+[_closeline(l) for l in [l19, l18, l17, l16, l15, lmean]]
+
+
+ax.set_xticks(theta)
+ax.set_xticklabels(deaths_headlines_w.index)
+plt.legend()
+plt.title("Deaths by week over years, Wales")
+plt.savefig('deaths-radar_wales.png')
+plt.show()
+```
+
+```python Collapsed="false"
+# Radar plot code taken from example at https://stackoverflow.com/questions/42878485/getting-matplotlib-radar-plot-with-pandas#
+
+fig = plt.figure(figsize=(10, 10))
+ax = fig.add_subplot(111, projection="polar")
+
+theta = np.roll(
+    np.flip(
+        np.arange(len(deaths_headlines_s))/float(len(deaths_headlines_s))*2.*np.pi),
+    14)
+l15, = ax.plot(theta, deaths_headlines_s[2015], color="#e47d7d", label="2015") # 0
+l16, = ax.plot(theta, deaths_headlines_s[2016], color="#afc169", label="2016") # 72 , d0e47d
+l17, = ax.plot(theta, deaths_headlines_s[2017], color="#7de4a6", label="2017") # 144
+l18, = ax.plot(theta, deaths_headlines_s[2018], color="#7da6e4", label="2018") # 216
+l19, = ax.plot(theta, deaths_headlines_s[2019], color="#d07de4", label="2019") # 288
+
+lmean, = ax.plot(theta, deaths_headlines_s['previous_mean'], color="black", linestyle='dashed', label="mean")
+
+l20, = ax.plot(theta, deaths_headlines_s[2020], color="red", label="2020")
+
+
+def _closeline(line):
+    x, y = line.get_data()
+    x = np.concatenate((x, [x[0]]))
+    y = np.concatenate((y, [y[0]]))
+    line.set_data(x, y)
+
+[_closeline(l) for l in [l19, l18, l17, l16, l15, lmean]]
+
+
+ax.set_xticks(theta)
+ax.set_xticklabels(deaths_headlines_s.index)
+plt.legend()
+plt.title("Deaths by week over years, Scotland")
+plt.savefig('deaths-radar_scotland.png')
+plt.show()
+```
+
+```python Collapsed="false"
+# Radar plot code taken from example at https://stackoverflow.com/questions/42878485/getting-matplotlib-radar-plot-with-pandas#
+
+fig = plt.figure(figsize=(10, 10))
+ax = fig.add_subplot(111, projection="polar")
+
+theta = np.roll(
+    np.flip(
+        np.arange(len(deaths_headlines_i))/float(len(deaths_headlines_i))*2.*np.pi),
+    14)
+l15, = ax.plot(theta, deaths_headlines_i[2015], color="#e47d7d", label="2015") # 0
+l16, = ax.plot(theta, deaths_headlines_i[2016], color="#afc169", label="2016") # 72 , d0e47d
+l17, = ax.plot(theta, deaths_headlines_i[2017], color="#7de4a6", label="2017") # 144
+l18, = ax.plot(theta, deaths_headlines_i[2018], color="#7da6e4", label="2018") # 216
+l19, = ax.plot(theta, deaths_headlines_i[2019], color="#d07de4", label="2019") # 288
+
+lmean, = ax.plot(theta, deaths_headlines_i['previous_mean'], color="black", linestyle='dashed', label="mean")
+
+l20, = ax.plot(theta, deaths_headlines_i[2020], color="red", label="2020")
+
+
+def _closeline(line):
+    x, y = line.get_data()
+    x = np.concatenate((x, [x[0]]))
+    y = np.concatenate((y, [y[0]]))
+    line.set_data(x, y)
+
+[_closeline(l) for l in [l19, l18, l17, l16, l15, lmean]]
+
+
+ax.set_xticks(theta)
+ax.set_xticklabels(deaths_headlines_i.index)
+plt.legend()
+plt.title("Deaths by week over years, Northern Ireland")
+plt.savefig('deaths-radar_northern_ireland.png')
+plt.show()
+```
+
+```python Collapsed="false"
+
+```
+
+```python Collapsed="false"
+
+```