In [2]:

import pandas as pd

John Hopkins csse_covid_19_daily_reports data inconsistent¶

https://github.com/CSSEGISandData/COVID-19/blob/master/csse_covid_19_data/csse_covid_19_daily_reports/02-04-2020.csv

versus

https://github.com/CSSEGISandData/COVID-19/blob/master/csse_covid_19_data/csse_covid_19_daily_reports/03-29-2020.csv

Lets assume Country_Region,Last_Update,Lat,Long_,Confirmed,Deaths,Recovered,Active

In [3]:

sg = pd.read_csv("singapore.csv", index_col=False, parse_dates=["Last_Update"], names=["Country_Region", "Last_Update", "Lat", "Long", "Confirmed", "Deaths","Recovered","Active"])

In [4]:

sg.shape

Out[4]:

(195, 8)

In [5]:

sg.head(3)

Out[5]:

	Country_Region	Last_Update	Lat	Deaths	Recovered	Active
0	Singapore	2020-02-02 01:03:32	18.0	NaN	NaN	NaN
1	Singapore	2020-02-02 01:03:32	18.0	NaN	NaN	NaN
2	Singapore	2020-02-04 15:33:03	24.0	NaN	NaN	NaN

In [6]:

sg.Deaths

Out[6]:

0       NaN
1       NaN
2       NaN
3       NaN
4       NaN
       ... 
190    27.0
191    27.0
192    27.0
193    27.0
194    27.0
Name: Deaths, Length: 195, dtype: float64

In [7]:

sg.dtypes

Out[7]:

Country_Region            object
Last_Update       datetime64[ns]
Lat                      float64
Long                     float64
Confirmed                  int64
Deaths                   float64
Recovered                float64
Active                   float64
dtype: object

In [26]:

import matplotlib.pyplot as plt
import mplcursors

---------------------------------------------------------------------------
ModuleNotFoundError                       Traceback (most recent call last)
<ipython-input-26-ed6bf68722a2> in <module>
      1 import matplotlib.pyplot as plt
----> 2 import mplcursors

ModuleNotFoundError: No module named 'mplcursors'

Howto plot with mathplot lib¶

https://discourse.matplotlib.org/t/gnuplot-to-matplotlib-time-series/21496/2

In [9]:

plt.plot("Last_Update", "Active", data=sg)

Out[9]:

[<matplotlib.lines.Line2D at 0x7f06766c9b20>]

Lag window¶

https://www.reddit.com/r/datascience/comments/ifhpiv/cumulative_to_daily_count_in_a_panda/g2nr9s0/?context=3

In [10]:

sg

Out[10]:

	Country_Region	Last_Update	Lat	Long	Confirmed	Deaths	Recovered	Active
0	Singapore	2020-02-02 01:03:32	18.0000	0.0000	0	NaN	NaN	NaN
1	Singapore	2020-02-02 01:03:32	18.0000	0.0000	0	NaN	NaN	NaN
2	Singapore	2020-02-04 15:33:03	24.0000	0.0000	0	NaN	NaN	NaN
3	Singapore	2020-02-05 16:33:03	28.0000	0.0000	0	NaN	NaN	NaN
4	Singapore	2020-02-05 16:33:03	28.0000	0.0000	0	NaN	NaN	NaN
...	...	...	...	...	...	...	...	...
190	Singapore	2020-08-18 04:27:56	1.2833	103.8333	55838	27.0	52350.0	3461.0
191	Singapore	2020-08-20 04:27:43	1.2833	103.8333	56031	27.0	52810.0	3194.0
192	Singapore	2020-08-21 04:27:41	1.2833	103.8333	56099	27.0	53119.0	2953.0
193	Singapore	2020-08-22 04:27:49	1.2833	103.8333	56216	27.0	53651.0	2538.0
194	Singapore	2020-08-23 04:27:48	1.2833	103.8333	56266	27.0	53920.0	2319.0

195 rows × 8 columns

In [12]:

sg['confirmed_day_before'] =  sg['Confirmed'].shift(1)  

In [14]:

sg['daily_increase'] = sg['Confirmed'] - sg['confirmed_day_before']    

In [15]:

sg

Out[15]:

	Country_Region	Last_Update	Lat	Long	Confirmed	Deaths	Recovered	Active	confirmed_day_before	daily_increase
0	Singapore	2020-02-02 01:03:32	18.0000	0.0000	0	NaN	NaN	NaN	NaN	NaN
1	Singapore	2020-02-02 01:03:32	18.0000	0.0000	0	NaN	NaN	NaN	0.0	0.0
2	Singapore	2020-02-04 15:33:03	24.0000	0.0000	0	NaN	NaN	NaN	0.0	0.0
3	Singapore	2020-02-05 16:33:03	28.0000	0.0000	0	NaN	NaN	NaN	0.0	0.0
4	Singapore	2020-02-05 16:33:03	28.0000	0.0000	0	NaN	NaN	NaN	0.0	0.0
...	...	...	...	...	...	...	...	...	...	...
190	Singapore	2020-08-18 04:27:56	1.2833	103.8333	55838	27.0	52350.0	3461.0	55747.0	91.0
191	Singapore	2020-08-20 04:27:43	1.2833	103.8333	56031	27.0	52810.0	3194.0	55838.0	193.0
192	Singapore	2020-08-21 04:27:41	1.2833	103.8333	56099	27.0	53119.0	2953.0	56031.0	68.0
193	Singapore	2020-08-22 04:27:49	1.2833	103.8333	56216	27.0	53651.0	2538.0	56099.0	117.0
194	Singapore	2020-08-23 04:27:48	1.2833	103.8333	56266	27.0	53920.0	2319.0	56216.0	50.0

195 rows × 10 columns

In [18]:

plt.plot("Last_Update", "daily_increase", data=sg)

Out[18]:

[<matplotlib.lines.Line2D at 0x7f067622dfa0>]

In [19]:

sg.drop_duplicates(subset=sg.columns, inplace=True)

In [20]:

plt.plot("Last_Update", "daily_increase", data=sg)

Out[20]:

[<matplotlib.lines.Line2D at 0x7f06761eca90>]

In [24]:

sg.to_excel("output.xlsx", sheet_name='covidsg')