test_and_case_data-be.md

   1 ---
   2 jupyter:
   3   jupytext:
   4     formats: ipynb,md
   5     text_representation:
   6       extension: .md
   7       format_name: markdown
   8       format_version: '1.3'
   9       jupytext_version: 1.10.2
  10   kernelspec:
  11     display_name: Python 3
  12     language: python
  13     name: python3
  14 ---
  15
  16 ```python Collapsed="false"
  17 import itertools
  18 import collections
  19 import json
  20 import pandas as pd
  21 import numpy as np
  22 from scipy.stats import gmean
  23 import datetime
  24
  25 import matplotlib as mpl
  26 import matplotlib.pyplot as plt
  27 import matplotlib.animation as ani
  28 %matplotlib inline
  29 ```
  30
  31 <!-- #region Collapsed="false" -->
  32 Belgian data from https://epistat.wiv-isp.be/covid/
  33 <!-- #endregion -->
  34
  35 ```python Collapsed="false"
  36 !curl "https://epistat.sciensano.be/Data/COVID19BE_tests.csv" > COVID19BE_tests.csv
  37 ```
  38
  39 ```python Collapsed="false"
  40 raw_data = pd.read_csv('COVID19BE_tests.csv',
  41                        parse_dates=[0], dayfirst=True,
  42                        keep_default_na=False, na_values = ['']
  43                       )
  44 ```
  45
  46 ```python Collapsed="false"
  47 raw_data.dtypes
  48 ```
  49
  50 ```python Collapsed="false"
  51 tests_data = raw_data.set_index('DATE').sort_index().groupby('DATE').sum()[:-1]
  52 tests_data
  53 ```
  54
  55 ```python Collapsed="false"
  56 tests_data.plot()
  57 ```
  58
  59 ```python Collapsed="false"
  60 # data_by_day.newAdmissions.dropna()
  61 ```
  62
  63 ```python Collapsed="false"
  64 data_by_day = pd.read_csv('data_by_day_be.csv', index_col='dateRep', parse_dates=True)
  65 data_by_day
  66 ```
  67
  68 ```python Collapsed="false"
  69 data_by_day.loc['2020-07-03']
  70 ```
  71
  72 ```python Collapsed="false"
  73 data_by_day = data_by_day.merge(tests_data[['TESTS_ALL']], how='outer',
  74     left_index=True, right_index=True).dropna()
  75 ```
  76
  77 ```python Collapsed="false"
  78 data_by_day
  79 ```
  80
  81 ```python Collapsed="false"
  82 data_by_day['deaths_m7'] = data_by_day.deaths.transform(lambda x: x.rolling(7, 1).mean())
  83 data_by_day['cases_m7'] = data_by_day.cases.transform(lambda x: x.rolling(7, 1).mean())
  84 data_by_day['tests_m7'] = data_by_day.TESTS_ALL.transform(lambda x: x.rolling(7, 1).mean())
  85 ```
  86
  87 ```python Collapsed="false"
  88 data_by_day = data_by_day.dropna()
  89 data_by_day
  90 ```
  91
  92 ```python Collapsed="false"
  93 data_by_day.loc['2020-06-22']
  94 ```
  95
  96 ```python Collapsed="false"
  97 data_by_day.loc['2020-04-15':'2020-08-26', ['cases_m7', 'tests_m7']].plot()
  98 ```
  99
 100 ```python Collapsed="false"
 101 data_by_day['fraction_positive'] = data_by_day.cases / data_by_day.TESTS_ALL
 102 data_by_day['fraction_positive_m7'] = data_by_day.cases_m7 / data_by_day.tests_m7
 103 ```
 104
 105 ```python Collapsed="false"
 106 data_by_day[['fraction_positive', 'fraction_positive_m7']].dropna().plot()
 107 ```
 108
 109 ```python Collapsed="false"
 110 # ax = data_by_day.dropna().loc['2020-06-15': , ['fraction_positive', 'fraction_positive_m7']].plot(figsize=(10, 8), title='Fraction of tests with positive results')
 111 # ax.legend(['Fraction positive per day', 'Fraction positive, 7 day moving average'])
 112 # ax.set_ylabel('Fraction positive')
 113 # plt.savefig('fraction_positive_tests.png')
 114 ```
 115
 116 ```python Collapsed="false"
 117 pri_y_max = int((data_by_day.dropna().loc['2020-06-15': , 'tests_m7'].max() * 1.1) / 100 ) * 100
 118 ax = data_by_day.dropna().loc['2020-06-15': , 'tests_m7'].plot(figsize=(10, 8),
 119                                                                style=['k-'],
 120                                                                legend=False,
 121                                                                ylim=(0, pri_y_max))
 122 ax.set_title('Tests done and new cases (7 day moving average)')
 123 ax.legend(['Tests, 7 day moving average'], loc='lower left')
 124 ax.set_ylabel('Tests')
 125 sec_y_max = int((data_by_day.dropna().loc['2020-06-15':, 'cases_m7'].max() * 1.1) / 100) * 100
 126 ax = data_by_day.dropna().loc['2020-06-15':, 'cases_m7'].plot(ax=ax, secondary_y=True, style='r--')
 127 ax.set_ylim((0, sec_y_max))
 128 ax.legend(['Cases (7 day moving average)'], loc='lower right')
 129 ax.set_ylabel('New cases')
 130 plt.savefig('tests_and_cases_be.png')
 131 ```
 132
 133 ```python Collapsed="false"
 134 int((sec_y_max * 1.1) / 100) * 100
 135 ```
 136
 137 ```python Collapsed="false"
 138 ax = (data_by_day.dropna().loc['2020-06-15': , ['fraction_positive', 'fraction_positive_m7']] * 100).plot(figsize=(10, 8),
 139                                                                                                   style=['b:', 'k-'], legend=False)
 140 ax.set_title('Fraction of tests with positive results')
 141 ax.legend(['Fraction positive (%)', 'Fraction positive (%), 7 day moving average'], loc='lower left')
 142 ax.set_ylabel('Fraction positive')
 143 ax = data_by_day.dropna().loc['2020-06-15':, 'cases_m7'].plot(ax=ax, secondary_y=True, style='r--')
 144 ax.legend(['Cases (7 day moving average)'], loc='lower right')
 145 ax.set_ylabel('New cases')
 146 plt.savefig('fraction_positive_tests_be.png')
 147 ```
 148
 149 ```python Collapsed="false"
 150 ax = data_by_day.dropna().loc['2020-06-15':].plot(x='fraction_positive_m7', y='tests_m7',
 151                                                   figsize=(8, 8),
 152                                                   legend=None)
 153 ax.set_xlabel("Fraction of tests that are positive")
 154 ax.set_ylabel("Number of tests")
 155 for d in data_by_day.dropna().loc['2020-06-15'::15].index:
 156     ax.plot(data_by_day.loc[d, 'fraction_positive_m7'], data_by_day.loc[d, 'tests_m7'], 'o',
 157                         markersize=8)#, markerfacecolor=marker_col, markeredgecolor=marker_col)
 158     ax.text(data_by_day.loc[d, 'fraction_positive_m7'] + 0.0002, data_by_day.loc[d, 'tests_m7'],
 159             s = d.strftime("%d %B %Y"))
 160 plt.savefig('fraction_positive_tests_vs_tests_be.png')
 161 ```
 162
 163 ```python Collapsed="false"
 164 data_by_day.dropna().loc['2020-06-15':][:2]
 165 ```
 166
 167 ```python Collapsed="false"
 168 fig = plt.figure(figsize=(8, 8))
 169 plt.ylabel('Number of tests')
 170 plt.xlabel('Fraction of tests that are positive')
 171 all_data = data_by_day.dropna().loc['2020-06-15':]
 172
 173
 174 minx = all_data.fraction_positive_m7.min() * 0.9
 175 maxx = all_data.fraction_positive_m7.max() * 1.1
 176 miny = all_data.tests_m7.min() * 0.9
 177 maxy = all_data.tests_m7.max() * 1.1
 178
 179 plt.xlim(minx, maxx)
 180 plt.ylim(miny, maxy)
 181 # plt.legend(None)
 182
 183 def build_state_frame(i):
 184     this_data = all_data[:i]
 185     plt.clf()
 186     plt.ylabel('Number of tests')
 187     plt.xlabel('Fraction of tests that are positive')
 188     plt.xlim(minx, maxx)
 189     plt.ylim(miny, maxy)
 190     p = plt.plot(this_data.fraction_positive_m7, this_data.tests_m7)
 191     p[0].set_color('r')
 192     for d in this_data[::15].index:
 193         plt.plot(this_data.loc[d, 'fraction_positive_m7'],
 194                 this_data.loc[d, 'tests_m7'], 'o',
 195                 markersize=8, markerfacecolor='r', markeredgecolor='r')
 196         plt.text(this_data.loc[d, 'fraction_positive_m7'] + 0.0002,
 197                 this_data.loc[d, 'tests_m7'],
 198             s = d.strftime("%d %B %Y"))
 199
 200 animator = ani.FuncAnimation(fig, build_state_frame,
 201                              frames=all_data.shape[0]+1,
 202                              interval=100,
 203                              repeat_delay=200,
 204                              repeat=True)
 205 animator.save('tests_vs_fraction_positive_be.mp4')
 206 plt.show()
 207 ```
 208
 209 ```python Collapsed="false"
 210 !rm tests_vs_fraction_positive_animation_be.png
 211 !ffmpeg -i tests_vs_fraction_positive_be.mp4 -plays 0 -final_delay 1 -f apng tests_vs_fraction_positive_animation_be.png
 212 ```
 213
 214 ```python Collapsed="false"
 215
 216 ```