| Download

Data from https://www.kaggle.com/sudalairajkumar/novel-corona-virus-2019-dataset

Project: tests and ideas

Path: coronavirus / coronavirus.ipynb

Views: ¹⁴¹¹
License: APACHE

Kernel: Python 3 (system-wide)

SARS-CoV-19 Corona Virus Data

Working with some datasets from Kaggle on CoCalc

https://www.kaggle.com/sudalairajkumar/novel-corona-virus-2019-dataset

Download covid_19_data.csv

Better: adjust all below to work with https://github.com/CSSEGISandData/COVID-19/blob/master/csse_covid_19_data/csse_covid_19_daily_reports/

In [1]:

import numpy as np
import pandas as pd
import matplotlib.pyplot as plt
import seaborn as sbn
sbn.set('notebook')

In [2]:

covid = pd.read_csv("covid_19_data.csv", parse_dates=[1, 4]).rename(
    columns={
        'Province/State': 'state',
        'Country/Region': 'country',
        'Last Update': 'lastupdate',
        'ObservationDate': 'time'
    })
covid.tail()

	SNo	time	state	country	lastupdate
4242	4243	2020-03-08	Northern Territory	Australia	2020-03-06 04:33:03
4243	4244	2020-03-08	Lackland, TX (From Diamond Princess)	US	2020-02-24 23:33:02
4244	4245	2020-03-08	Montgomery County, TX	US	2020-03-07 19:53:02
4245	4246	2020-03-08	Omaha, NE (From Diamond Princess)	US	2020-02-24 23:33:02
4246	4247	2020-03-08	Travis, CA (From Diamond Princess)	US	2020-02-24 23:33:02

In [3]:

covid['state'] = covid['state'].astype('category')
covid['country'] = covid['country'].astype('category')

In [4]:

covid.dtypes

SNo                    int64
time          datetime64[ns]
state               category
country             category
lastupdate    datetime64[ns]
Confirmed            float64
Deaths               float64
Recovered            float64
dtype: object

"sick" are those, who are neither dead nor are recovered

In [5]:

covid['sick'] = covid.Confirmed - covid.Recovered - covid.Deaths

"deathrate" based on confirmed cases

In [6]:

covid['deathrate'] = 100 * covid.Deaths / covid.Confirmed
covid['recrate'] = 100 * covid.Recovered / covid.Confirmed

filter latest data only!

In [7]:

# but keep the full data around …
covid_all = covid.copy()

In [8]:

covid = covid[covid.time == covid.time.max()].reindex()
covid

	SNo	time	state	country	lastupdate	Confirmed	Deaths	Recovered	sick	deathrate	recrate
3992	3993	2020-03-08	Hubei	Mainland China	2020-03-08 14:43:03	67707.0	2986.0	45235.0	19486.0	4.410179	66.809931
3993	3994	2020-03-08	NaN	Italy	2020-03-08 18:03:04	7375.0	366.0	622.0	6387.0	4.962712	8.433898
3994	3995	2020-03-08	NaN	South Korea	2020-03-08 12:53:03	7314.0	50.0	118.0	7146.0	0.683620	1.613344
3995	3996	2020-03-08	NaN	Iran	2020-03-08 11:03:30	6566.0	194.0	2134.0	4238.0	2.954615	32.500761
3996	3997	2020-03-08	Guangdong	Mainland China	2020-03-08 14:43:03	1352.0	7.0	1256.0	89.0	0.517751	92.899408
...	...	...	...	...	...	...	...	...	...	...	...
4242	4243	2020-03-08	Northern Territory	Australia	2020-03-06 04:33:03	0.0	0.0	0.0	0.0	NaN	NaN
4243	4244	2020-03-08	Lackland, TX (From Diamond Princess)	US	2020-02-24 23:33:02	0.0	0.0	0.0	0.0	NaN	NaN
4244	4245	2020-03-08	Montgomery County, TX	US	2020-03-07 19:53:02	0.0	0.0	0.0	0.0	NaN	NaN
4245	4246	2020-03-08	Omaha, NE (From Diamond Princess)	US	2020-02-24 23:33:02	0.0	0.0	0.0	0.0	NaN	NaN
4246	4247	2020-03-08	Travis, CA (From Diamond Princess)	US	2020-02-24 23:33:02	0.0	0.0	0.0	0.0	NaN	NaN

255 rows × 11 columns

In [9]:

covid.describe()

	SNo	Confirmed	Deaths	Recovered	sick	deathrate	recrate
count	255.000000	255.00000	255.000000	255.000000	255.000000	249.000000	249.000000
mean	4120.000000	430.72549	14.913725	238.019608	177.792157	1.320539	16.819022
std	73.756356	4302.71234	188.609605	2838.177099	1382.094730	7.645967	33.098950
min	3993.000000	0.00000	0.000000	0.000000	0.000000	0.000000	0.000000
25%	4056.500000	1.00000	0.000000	0.000000	1.000000	0.000000	0.000000
50%	4120.000000	4.00000	0.000000	0.000000	3.000000	0.000000	0.000000
75%	4183.500000	35.50000	0.000000	1.000000	16.500000	0.000000	10.000000
max	4247.000000	67707.00000	2986.000000	45235.000000	19486.000000	100.000000	100.000000

death rate stats, where at least a significant amount of patients are known …

In [10]:

min_deaths = covid[covid.Deaths > 2].Deaths.quantile(.1)
min_deaths

3.0

In [11]:

deathrate_sign = covid[covid.Deaths > min_deaths].deathrate
deathrate_sign.describe()

count    20.000000
mean      3.243667
std       4.624665
min       0.392927
25%       0.844441
50%       1.799359
75%       3.108818
max      20.481928
Name: deathrate, dtype: float64

In [12]:

plt.plot(np.sort(deathrate_sign.to_numpy()))

[<matplotlib.lines.Line2D at 0x7f09f9525240>]

In [0]:

In [13]:

covid.sort_values(by='Deaths', ascending=False).head(20)

	SNo	time	state	country	lastupdate	Confirmed	Deaths	Recovered	sick	deathrate	recrate
3992	3993	2020-03-08	Hubei	Mainland China	2020-03-08 14:43:03	67707.0	2986.0	45235.0	19486.0	4.410179	66.809931
3993	3994	2020-03-08	NaN	Italy	2020-03-08 18:03:04	7375.0	366.0	622.0	6387.0	4.962712	8.433898
3995	3996	2020-03-08	NaN	Iran	2020-03-08 11:03:30	6566.0	194.0	2134.0	4238.0	2.954615	32.500761
3994	3995	2020-03-08	NaN	South Korea	2020-03-08 12:53:03	7314.0	50.0	118.0	7146.0	0.683620	1.613344
3997	3998	2020-03-08	Henan	Mainland China	2020-03-08 05:03:02	1272.0	22.0	1247.0	3.0	1.729560	98.034591
3999	4000	2020-03-08	NaN	France	2020-03-08 18:03:04	1126.0	19.0	12.0	1095.0	1.687389	1.065719
4037	4038	2020-03-08	King County, WA	US	2020-03-08 20:23:09	83.0	17.0	1.0	65.0	20.481928	1.204819
4006	4007	2020-03-08	NaN	Spain	2020-03-08 20:33:02	673.0	17.0	30.0	626.0	2.526003	4.457652
4011	4012	2020-03-08	Heilongjiang	Mainland China	2020-03-08 14:43:03	481.0	13.0	412.0	56.0	2.702703	85.654886
4012	4013	2020-03-08	Beijing	Mainland China	2020-03-08 01:23:07	428.0	8.0	308.0	112.0	1.869159	71.962617
3996	3997	2020-03-08	Guangdong	Mainland China	2020-03-08 14:43:03	1352.0	7.0	1256.0	89.0	0.517751	92.899408
4008	4009	2020-03-08	Chongqing	Mainland China	2020-03-08 23:23:03	576.0	6.0	527.0	43.0	1.041667	91.493056
4015	4016	2020-03-08	Hebei	Mainland China	2020-03-07 13:03:05	318.0	6.0	307.0	5.0	1.886792	96.540881
4025	4026	2020-03-08	Hainan	Mainland China	2020-03-08 04:33:02	168.0	6.0	159.0	3.0	3.571429	94.642857
4010	4011	2020-03-08	NaN	Japan	2020-03-08 14:53:11	502.0	6.0	76.0	420.0	1.195219	15.139442
4004	4005	2020-03-08	Shandong	Mainland China	2020-03-08 10:03:11	758.0	6.0	642.0	110.0	0.791557	84.696570
4005	4006	2020-03-08	Diamond Princess cruise ship	Others	2020-03-06 01:29:39	696.0	6.0	40.0	650.0	0.862069	5.747126
4044	4045	2020-03-08	NaN	Iraq	2020-03-08 21:03:03	60.0	6.0	0.0	54.0	10.000000	0.000000
4002	4003	2020-03-08	Anhui	Mainland China	2020-03-08 05:13:06	990.0	6.0	984.0	0.0	0.606061	99.393939
4001	4002	2020-03-08	Hunan	Mainland China	2020-03-08 12:53:03	1018.0	4.0	968.0	46.0	0.392927	95.088409

Situation in Italy

In [14]:

italy = covid_all[covid_all.country == 'Italy'].sort_values(by='lastupdate')
italy.plot.line(x='lastupdate', y=['Confirmed', 'Deaths', 'Recovered', 'sick'])

<matplotlib.axes._subplots.AxesSubplot at 0x7f09f90971d0>

In [15]:

italy.plot.line(x='lastupdate', y='deathrate')

<matplotlib.axes._subplots.AxesSubplot at 0x7f09f8fc2828>

In [22]:

italy.plot.line(x='lastupdate', y='recrate')

<matplotlib.axes._subplots.AxesSubplot at 0x7f09f8f066a0>

In [0]:

outside china

In [16]:

outside_ch = covid[covid.country != 'Mainland China']
outside_ch

	SNo	time	state	country	lastupdate	Confirmed	Deaths	Recovered	sick	deathrate	recrate
3993	3994	2020-03-08	NaN	Italy	2020-03-08 18:03:04	7375.0	366.0	622.0	6387.0	4.962712	8.433898
3994	3995	2020-03-08	NaN	South Korea	2020-03-08 12:53:03	7314.0	50.0	118.0	7146.0	0.683620	1.613344
3995	3996	2020-03-08	NaN	Iran	2020-03-08 11:03:30	6566.0	194.0	2134.0	4238.0	2.954615	32.500761
3999	4000	2020-03-08	NaN	France	2020-03-08 18:03:04	1126.0	19.0	12.0	1095.0	1.687389	1.065719
4000	4001	2020-03-08	NaN	Germany	2020-03-08 21:03:03	1040.0	0.0	18.0	1022.0	0.000000	1.730769
...	...	...	...	...	...	...	...	...	...	...	...
4242	4243	2020-03-08	Northern Territory	Australia	2020-03-06 04:33:03	0.0	0.0	0.0	0.0	NaN	NaN
4243	4244	2020-03-08	Lackland, TX (From Diamond Princess)	US	2020-02-24 23:33:02	0.0	0.0	0.0	0.0	NaN	NaN
4244	4245	2020-03-08	Montgomery County, TX	US	2020-03-07 19:53:02	0.0	0.0	0.0	0.0	NaN	NaN
4245	4246	2020-03-08	Omaha, NE (From Diamond Princess)	US	2020-02-24 23:33:02	0.0	0.0	0.0	0.0	NaN	NaN
4246	4247	2020-03-08	Travis, CA (From Diamond Princess)	US	2020-02-24 23:33:02	0.0	0.0	0.0	0.0	NaN	NaN

224 rows × 11 columns

In [17]:

outside_ch.pivot_table(index='country', values=['Deaths', 'Confirmed', 'sick', 'Recovered'], aggfunc=np.sum).sort_values(by='Deaths', ascending=False).head(20)

	Confirmed	Deaths	Recovered	sick
country
Italy	7375.0	366.0	622.0	6387.0
Iran	6566.0	194.0	2134.0	4238.0
South Korea	7314.0	50.0	118.0	7146.0
US	537.0	21.0	8.0	508.0
France	1126.0	19.0	12.0	1095.0
Spain	673.0	17.0	30.0	626.0
Iraq	60.0	6.0	0.0	54.0
Japan	502.0	6.0	76.0	420.0
Others	696.0	6.0	40.0	650.0
Australia	76.0	4.0	21.0	51.0
Hong Kong	114.0	3.0	58.0	53.0
UK	273.0	3.0	18.0	252.0
Netherlands	265.0	3.0	0.0	262.0
Switzerland	337.0	2.0	3.0	332.0
Philippines	10.0	1.0	1.0	8.0
Egypt	49.0	1.0	1.0	47.0
Taiwan	45.0	1.0	13.0	31.0
Thailand	50.0	1.0	31.0	18.0
Argentina	12.0	1.0	0.0	11.0
San Marino	36.0	1.0	0.0	35.0

In [18]:

covid_all['inchina'] = covid_all.country == 'Mainland China'

In [19]:

covid_all = covid_all.sort_values('lastupdate')

In [28]:

covid_all_inchina = covid_all[covid_all.inchina == False].dropna().pivot_table(index='lastupdate', values=['Deaths', 'Recovered', 'sick', 'Confirmed'], aggfunc=np.sum)
covid_all_inchina

	Confirmed	Deaths	Recovered	sick
lastupdate
2020-01-22 17:00:00	3.0	0.0	0.0	3.0
2020-01-23 17:00:00	6.0	0.0	0.0	6.0
2020-01-24 17:00:00	9.0	0.0	0.0	9.0
2020-01-25 17:00:00	12.0	0.0	0.0	12.0
2020-01-26 16:00:00	23.0	0.0	0.0	23.0
...	...	...	...	...
2020-03-08 21:23:03	44.0	0.0	1.0	43.0
2020-03-08 21:33:02	30.0	0.0	3.0	27.0
2020-03-08 21:43:03	55.0	1.0	0.0	54.0
2020-03-08 21:53:03	1.0	0.0	0.0	1.0
2020-03-08 21:53:04	5.0	0.0	0.0	5.0

294 rows × 4 columns

In [29]:

covid_all_inchina.dropna().sick.plot.line()

<matplotlib.axes._subplots.AxesSubplot at 0x7f09f8da3208>

In [0]:

In [0]: