[Python] Using Python for data analysis--a case study of a new coronavirus epidemic

Important Notes

Submit only this document, and then submit it it to the "homework after completion" of the course.

1. File Naming:

Must be in number-name-major-class. Naming of ipynb, (e.g. 202011030101-Qiao Feng-Measurement and Control Technology and Instrument-Measurement and Control 20-1.ipynb)

In addition, the second line at the beginning of the document (marked in the document) is changed to your own number-name-major-class

2. Deadline:

24:00 p.m. on April 30, 2021

3. Scoring rules:

Analysis Document: Completeness: Code Quality 3:5:2

Analytical documents refer to the ideas, explanations, and explanations of the results of your analysis of data (short and concise, not for writing).

ps: You write your own code far better than anything else, no matter how beautiful or ugly it is, just ask today is better than yesterday! Come on.

Reminder:

The epidemic is still ravaging, please actively protect yourself

I wish you all good results

Since there is too much data, try using head() or tail() to view the data so that the program does not respond for a long time

=======================

Data for this project come from Syringa orchard. The main purpose of this project is to better understand the epidemic situation and its development through the analysis of historical epidemic data, and to provide data support for decision-making on fighting the epidemic situation.

1. Ask questions

From the national level, your provinces and cities, the epidemic situation abroad and other three aspects, the main research issues are as follows:

(1) What is the trend of cumulative diagnosis/suspicion/cure/death over time throughout the country?

(2) What is the situation in your province or city?

(3) What is the overall situation of the global epidemic?

(4) Give your judgment on the epidemic trend in the next half year based on your analysis results, and what suggestions do you have for individuals and society in fighting the epidemic?

2. Understanding data

Original dataset: AreaInfo.csv, import related packages, read data, and assign areas

import pandas as pd
import numpy as np
import matplotlib.pyplot as plt

areas = pd.read_csv('data\AreaInfo.csv')
areas

	continentName	continentEnglishName	countryName	countryEnglishName	provinceName	provinceEnglishName	province_zipCode	province_confirmedCount	province_suspectedCount	province_curedCount	province_deadCount	updateTime	cityName	cityEnglishName	city_zipCode	city_confirmedCount	city_suspectedCount	city_curedCount	city_deadCount
0	Asia	Asia	China	China	Macao	Macau	820000	47	9.0	46	0	2021-01-22 23:40:08	NaN	NaN	NaN	NaN	NaN	NaN	NaN
1	North America	North America	U.S.A	United States of America	U.S.A	United States of America	971002	24632468	0.0	10845438	410378	2021-01-22 23:40:08	NaN	NaN	NaN	NaN	NaN	NaN	NaN
2	South America	South America	Brazil	Brazil	Brazil	Brazil	973003	8699814	0.0	7580741	214228	2021-01-22 23:40:08	NaN	NaN	NaN	NaN	NaN	NaN	NaN
3	Europe	Europe	Belgium	Belgium	Belgium	Belgium	961001	686827	0.0	19239	20620	2021-01-22 23:40:08	NaN	NaN	NaN	NaN	NaN	NaN	NaN
4	Europe	Europe	Russia	Russia	Russia	Russia	964006	3677352	0.0	3081536	68412	2021-01-22 23:40:08	NaN	NaN	NaN	NaN	NaN	NaN	NaN
...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...
429911	Asia	Asia	China	China	Liaoning Province	Liaoning	210000	0	1.0	0	0	2020-01-22 03:28:10	NaN	NaN	NaN	NaN	NaN	NaN	NaN
429912	Asia	Asia	China	China	Taiwan	Taiwan	710000	1	0.0	0	0	2020-01-22 03:28:10	NaN	NaN	NaN	NaN	NaN	NaN	NaN
429913	Asia	Asia	China	Hongkong	Hong Kong	Hongkong	810000	0	117.0	0	0	2020-01-22 03:28:10	NaN	NaN	NaN	NaN	NaN	NaN	NaN
429914	Asia	Asia	China	China	Heilongjiang Province	Heilongjiang	230000	0	1.0	0	0	2020-01-22 03:28:10	NaN	NaN	NaN	NaN	NaN	NaN	NaN
429915	Asia	Asia	China	China	Hunan Province	Hunan	430000	1	0.0	0	0	2020-01-22 03:28:10	NaN	NaN	NaN	NaN	NaN	NaN	NaN

429916 rows × 19 columns

areas.info()

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 429916 entries, 0 to 429915
Data columns (total 19 columns):
 #   Column                   Non-Null Count   Dtype  
---  ------                   --------------   -----  
 0   continentName            429872 non-null  object 
 1   continentEnglishName     429872 non-null  object 
 2   countryName              429916 non-null  object 
 3   countryEnglishName       403621 non-null  object 
 4   provinceName             429916 non-null  object 
 5   provinceEnglishName      403621 non-null  object 
 6   province_zipCode         429916 non-null  int64  
 7   province_confirmedCount  429916 non-null  int64  
 8   province_suspectedCount  429913 non-null  float64
 9   province_curedCount      429916 non-null  int64  
 10  province_deadCount       429916 non-null  int64  
 11  updateTime               429916 non-null  object 
 12  cityName                 125143 non-null  object 
 13  cityEnglishName          119660 non-null  object 
 14  city_zipCode             123877 non-null  float64
 15  city_confirmedCount      125143 non-null  float64
 16  city_suspectedCount      125143 non-null  float64
 17  city_curedCount          125143 non-null  float64
 18  city_deadCount           125143 non-null  float64
dtypes: float64(6), int64(4), object(9)
memory usage: 62.3+ MB

View and statistics to get a general idea of the data

Related field meaning introduction:

Tips:

The provinceName of foreign data is not a province name, but is labeled with its country name, that is, the data is no longer subdivided into provinces.

There is also a record like'China'in the provinceName of China data, which represents the total of the provinces in the country on that day. If you use it well, it will be much easier to analyze the national situation.

3. Data cleaning

(1) Basic data processing

Data cleaning mainly includes: selecting subsets, missing data processing, data format conversion, outlier data processing, etc.

Tip: Because the data is reported by all countries, it is inevitable to conceal the missing report when the situation is critical, so there are many missing values, which can be completed or discarded. See "Treatment of missing values in Pandas.ipynb"

Selection of domestic epidemic data (the final selected data is named china)

Select domestic epidemic data
For the updateTime column, convert it to a date type, extract the year-month-day, and view the processing results. (Tip: dt.date)
Since the data is updated hourly every day and there are many duplicates in a day, please go back and keep only the latest data in a day.

Tip: df.drop_duplicates(subset=['provinceName','updateTime'], keep='first', inplace=False)

Where df is the DataFrame of your choice for domestic epidemic data

Remove columns that are not in the scope of this study, leaving only ['continentName','countryName','provinceName','province_confirmedCount','province_suspectedCount','province_curedCount','province_deadCount','updateTime'] columns and index them with'updateTime'.

Tip: Both methods can: (1) Select these columns or (2) Remove the remaining columns

# The code is given here, and the acquisition of later provincial and global data is much the same.
china = areas.loc[areas.countryName=='China',:].copy()
china['updateTime'] = pd.to_datetime(china.updateTime,format="%Y-%m-%d",errors='coerce').dt.date
china = china.drop_duplicates(subset=['provinceName', 'updateTime'], keep='first', inplace=False)

# Convert Character Type Date Column (Index) to Datetime Index
china['updateTime'] = pd.to_datetime(china['updateTime'])
china.set_index('updateTime',inplace=True)

china = china[['continentName','countryName','provinceName','province_confirmedCount','province_suspectedCount','province_curedCount','province_deadCount']]
china = china[china.provinceName=='China']
china.head()

	continentName	countryName	provinceName	province_confirmedCount	province_suspectedCount	province_curedCount	province_deadCount
updateTime
2021-01-22	Asia	China	China	99667	0.0	92275	4810
2021-01-21	Asia	China	China	99513	0.0	92198	4809
2021-01-20	Asia	China	China	99285	0.0	92130	4808
2021-01-19	Asia	China	China	99094	0.0	92071	4806
2021-01-18	Asia	China	China	98922	0.0	91994	4805

Check the data information for missing data/data type correctness. If there are missing values, you can complete or discard them, see ** "Processing missing values for Pandas. ipynb"**

# Check the data information for missing data/data type correctness. If there are missing values, they can be completed or discarded
china.info()
china.head(5)

<class 'pandas.core.frame.DataFrame'>
DatetimeIndex: 305 entries, 2021-01-22 to 2020-03-15
Data columns (total 7 columns):
 #   Column                   Non-Null Count  Dtype  
---  ------                   --------------  -----  
 0   continentName            305 non-null    object 
 1   countryName              305 non-null    object 
 2   provinceName             305 non-null    object 
 3   province_confirmedCount  305 non-null    int64  
 4   province_suspectedCount  305 non-null    float64
 5   province_curedCount      305 non-null    int64  
 6   province_deadCount       305 non-null    int64  
dtypes: float64(1), int64(3), object(3)
memory usage: 19.1+ KB

	continentName	countryName	provinceName	province_confirmedCount	province_suspectedCount	province_curedCount	province_deadCount
updateTime
2021-01-22	Asia	China	China	99667	0.0	92275	4810
2021-01-21	Asia	China	China	99513	0.0	92198	4809
2021-01-20	Asia	China	China	99285	0.0	92130	4808
2021-01-19	Asia	China	China	99094	0.0	92071	4806
2021-01-18	Asia	China	China	98922	0.0	91994	4805

Selection of epidemic data in your province or city (the final selected data is named myhome)

This step can also be done later

Select epidemic data of the province or city (refined to city; if municipality directly under the Central Government, refined to district)
For the updateTime column, convert it to a date type, extract the year-month-day, and view the processing results. (Tip: dt.date)
Since the data is updated hourly every day and there are many duplicates in a day, please recreate and keep only the latest data in a day and index the rows with'updateTime'.

Tip: df.drop_duplicates(subset=['cityName','updateTime'], keep='first', inplace=False)

Remove columns not covered by this study

Tip: df.drop(['continentName','continentEnglishName','countryName','countryEnglishName','provinceEnglishName',
'province_zipCode','cityEnglishName','updateTime','city_zipCode'],axis=1,inplace=True)

Where df is the DataFrame for the epidemic data of the province or city you selected

# Provincial and municipal data acquisition
df = areas.loc[areas.provinceName=='Sichuan Province',:].copy()
df
df['updateTime'] = pd.to_datetime(df.updateTime,format="%Y-%m-%d",errors='coerce').dt.date
df = df.drop_duplicates(subset=['provinceName', 'updateTime'], keep='first', inplace=False)

# Convert Character Type Date Column (Index) to Datetime Index
df['updateTime'] = pd.to_datetime(df['updateTime'])
# df.set_index('updateTime',inplace=True)

df.drop(['continentName','continentEnglishName','countryName','countryEnglishName','provinceEnglishName','province_zipCode','cityEnglishName','updateTime','city_zipCode'],axis=1,inplace=True)
# df = df[df.provinceName=='Hubei Province']
df
# df.head(10)

	provinceName	province_confirmedCount	province_suspectedCount	province_curedCount	province_deadCount	cityName	city_confirmedCount	city_suspectedCount	city_curedCount	city_deadCount
2538	Sichuan Province	865	14.0	852	3	Chengdu	468.0	13.0	455.0	3.0
7192	Sichuan Province	865	14.0	852	3	Chengdu	468.0	13.0	455.0	3.0
8912	Sichuan Province	863	14.0	852	3	Chengdu	466.0	13.0	455.0	3.0
10704	Sichuan Province	863	14.0	847	3	Chengdu	466.0	13.0	450.0	3.0
12175	Sichuan Province	862	14.0	847	3	Chengdu	465.0	13.0	450.0	3.0
...	...	...	...	...	...	...	...	...	...	...
427258	Sichuan Province	44	0.0	0	0	Chengdu	22.0	0.0	0.0	0.0
428070	Sichuan Province	28	0.0	0	0	Chengdu	16.0	0.0	0.0	0.0
428837	Sichuan Province	15	0.0	0	0	Chengdu	7.0	0.0	0.0	0.0
429846	Sichuan Province	8	2.0	0	0	NaN	NaN	NaN	NaN	NaN
429866	Sichuan Province	5	2.0	0	0	NaN	NaN	NaN	NaN	NaN

267 rows × 10 columns

df.info()
df.head()

<class 'pandas.core.frame.DataFrame'>
Int64Index: 267 entries, 2538 to 429866
Data columns (total 10 columns):
 #   Column                   Non-Null Count  Dtype  
---  ------                   --------------  -----  
 0   provinceName             267 non-null    object 
 1   province_confirmedCount  267 non-null    int64  
 2   province_suspectedCount  267 non-null    float64
 3   province_curedCount      267 non-null    int64  
 4   province_deadCount       267 non-null    int64  
 5   cityName                 265 non-null    object 
 6   city_confirmedCount      265 non-null    float64
 7   city_suspectedCount      265 non-null    float64
 8   city_curedCount          265 non-null    float64
 9   city_deadCount           265 non-null    float64
dtypes: float64(5), int64(3), object(2)
memory usage: 22.9+ KB

	provinceName	province_confirmedCount	province_suspectedCount	province_curedCount	province_deadCount	cityName	city_confirmedCount	city_suspectedCount	city_curedCount	city_deadCount
2538	Sichuan Province	865	14.0	852	3	Chengdu	468.0	13.0	455.0	3.0
7192	Sichuan Province	865	14.0	852	3	Chengdu	468.0	13.0	455.0	3.0
8912	Sichuan Province	863	14.0	852	3	Chengdu	466.0	13.0	455.0	3.0
10704	Sichuan Province	863	14.0	847	3	Chengdu	466.0	13.0	450.0	3.0
12175	Sichuan Province	862	14.0	847	3	Chengdu	465.0	13.0	450.0	3.0

df_clean = df.dropna()
df_clean.info()

<class 'pandas.core.frame.DataFrame'>
Int64Index: 265 entries, 2538 to 428837
Data columns (total 10 columns):
 #   Column                   Non-Null Count  Dtype  
---  ------                   --------------  -----  
 0   provinceName             265 non-null    object 
 1   province_confirmedCount  265 non-null    int64  
 2   province_suspectedCount  265 non-null    float64
 3   province_curedCount      265 non-null    int64  
 4   province_deadCount       265 non-null    int64  
 5   cityName                 265 non-null    object 
 6   city_confirmedCount      265 non-null    float64
 7   city_suspectedCount      265 non-null    float64
 8   city_curedCount          265 non-null    float64
 9   city_deadCount           265 non-null    float64
dtypes: float64(5), int64(3), object(2)
memory usage: 22.8+ KB

Check the data information for missing data/data type correctness. If there are missing values, you can complete or discard them, see ** "Processing missing values for Pandas. ipynb"**

Global epidemic data selection (the final selected data is named world)

This step can also be done later

Selecting epidemic data from abroad

Tip: Select foreign epidemic data (countryName!='China')

For the updateTime column, convert it to a date type, extract the year-month-day, and view the processing results. (Tip: dt.date)
Since the data is updated hourly every day and there are many duplicates in a day, please go back and keep only the latest data in a day.

Tip: df.drop_duplicates(subset=['provinceName','updateTime'], keep='first', inplace=False)

Where df is the DataFrame of your choice for domestic epidemic data

Remove columns that are not in the scope of this study, leaving only ['continentName','countryName','provinceName','province_confirmedCount','province_suspectedCount','province_curedCount','province_deadCount','updateTime'] columns and index them with'updateTime'.

Tip: Both methods can: (1) Select these columns or (2) Remove the remaining columns

Get global data

Tip: Use the concat function to connect the previous china with foreign data by "axis" to get global data.

# Provincial and municipal data acquisition
o = areas.loc[areas.countryName!='China',:].copy()
# df = areas.loc[areas.provinceName=='Hubei Province',:]. copy()
o
o['updateTime'] = pd.to_datetime(o.updateTime,format="%Y-%m-%d",errors='coerce').dt.date
o = o.drop_duplicates(subset=['provinceName', 'updateTime'], keep='first', inplace=False)

# Convert Character Type Date Column (Index) to Datetime Index
o['updateTime'] = pd.to_datetime(o['updateTime'])
# o.set_index('updateTime',inplace=True)

o = o[['continentName','countryName','provinceName','province_confirmedCount','province_suspectedCount','province_curedCount','province_deadCount','updateTime']]

o.set_index('updateTime',inplace=True)
o

	continentName	countryName	provinceName	province_confirmedCount	province_suspectedCount	province_curedCount	province_deadCount
updateTime
2021-01-22	North America	U.S.A	U.S.A	24632468	0.0	10845438	410378
2021-01-22	South America	Brazil	Brazil	8699814	0.0	7580741	214228
2021-01-22	Europe	Belgium	Belgium	686827	0.0	19239	20620
2021-01-22	Europe	Russia	Russia	3677352	0.0	3081536	68412
2021-01-22	Europe	Serbia	Serbia	436121	0.0	50185	5263
...	...	...	...	...	...	...	...
2020-01-27	NaN	Malaysia	Malaysia	3	0.0	0	0
2020-01-27	NaN	France	France	3	0.0	0	0
2020-01-27	NaN	Vietnam?	Vietnam?	2	0.0	0	0
2020-01-27	NaN	Nepal	Nepal	1	0.0	0	0
2020-01-27	NaN	Canada	Canada	1	0.0	0	0

62110 rows × 7 columns

world = pd.concat([china,o])
world

	continentName	countryName	provinceName	province_confirmedCount	province_suspectedCount	province_curedCount	province_deadCount
updateTime
2021-01-22	Asia	China	China	99667	0.0	92275	4810
2021-01-21	Asia	China	China	99513	0.0	92198	4809
2021-01-20	Asia	China	China	99285	0.0	92130	4808
2021-01-19	Asia	China	China	99094	0.0	92071	4806
2021-01-18	Asia	China	China	98922	0.0	91994	4805
...	...	...	...	...	...	...	...
2020-01-27	NaN	Malaysia	Malaysia	3	0.0	0	0
2020-01-27	NaN	France	France	3	0.0	0	0
2020-01-27	NaN	Vietnam?	Vietnam?	2	0.0	0	0
2020-01-27	NaN	Nepal	Nepal	1	0.0	0	0
2020-01-27	NaN	Canada	Canada	1	0.0	0	0

62415 rows × 7 columns

world.info()
world_clean = world.dropna()
world_clean.info()

<class 'pandas.core.frame.DataFrame'>
DatetimeIndex: 62415 entries, 2021-01-22 to 2020-01-27
Data columns (total 7 columns):
 #   Column                   Non-Null Count  Dtype  
---  ------                   --------------  -----  
 0   continentName            62377 non-null  object 
 1   countryName              62415 non-null  object 
 2   provinceName             62415 non-null  object 
 3   province_confirmedCount  62415 non-null  int64  
 4   province_suspectedCount  62415 non-null  float64
 5   province_curedCount      62415 non-null  int64  
 6   province_deadCount       62415 non-null  int64  
dtypes: float64(1), int64(3), object(3)
memory usage: 3.8+ MB
<class 'pandas.core.frame.DataFrame'>
DatetimeIndex: 62377 entries, 2021-01-22 to 2020-01-29
Data columns (total 7 columns):
 #   Column                   Non-Null Count  Dtype  
---  ------                   --------------  -----  
 0   continentName            62377 non-null  object 
 1   countryName              62377 non-null  object 
 2   provinceName             62377 non-null  object 
 3   province_confirmedCount  62377 non-null  int64  
 4   province_suspectedCount  62377 non-null  float64
 5   province_curedCount      62377 non-null  int64  
 6   province_deadCount       62377 non-null  int64  
dtypes: float64(1), int64(3), object(3)
memory usage: 3.8+ MB

Check the data information for missing data/data type correctness.

Tip: There are many missing values because the data is reported by all countries and it is unavoidable to conceal the missing report when the situation is critical. You can complete or discard the missing values. See ** "Handling missing values in Pandas.ipynb"**

4. Data analysis and visualization

For data analysis and visualization, select the required variables based on each question and create a new DataFrame for analysis and visualization so that the data is less cluttered and more organized.

Basic analysis

For basic analysis, only numpy, pandas, and matplotlib libraries are allowed.

Multiple coordinate systems can be displayed on one graph or on multiple graphs

Please select the type of graphics (line chart, pie chart, histogram, scatter chart, etc.) according to the analysis purpose. It is really no idea to go to Baidu epidemic map or other sites for epidemic analysis to inspire.

(1) What is the trend of cumulative diagnosis/cure/death over time throughout the country?

china

	continentName	countryName	provinceName	province_confirmedCount	province_suspectedCount	province_curedCount	province_deadCount
updateTime
2021-01-22	Asia	China	China	99667	0.0	92275	4810
2021-01-21	Asia	China	China	99513	0.0	92198	4809
2021-01-20	Asia	China	China	99285	0.0	92130	4808
2021-01-19	Asia	China	China	99094	0.0	92071	4806
2021-01-18	Asia	China	China	98922	0.0	91994	4805
...	...	...	...	...	...	...	...
2020-03-19	Asia	China	China	81263	0.0	70561	3250
2020-03-18	Asia	China	China	81202	0.0	69777	3242
2020-03-17	Asia	China	China	81135	0.0	68820	3231
2020-03-16	Asia	China	China	81099	0.0	67930	3218
2020-03-15	Asia	China	China	81062	0.0	67037	3204

305 rows × 7 columns

# Accumulated diagnosis nationwide
plt.figure()
plt.plot(china.province_confirmedCount)
plt.title('China')
plt.xlabel('Time')
plt.ylabel('Count')
plt.show()

# Accumulated national cure
plt.figure()
plt.plot(china.province_curedCount)
plt.title('China')
plt.xlabel('Time')
plt.ylabel('Count')
plt.show()

# Accumulated national deaths
plt.figure()
plt.plot(china.province_deadCount)
plt.title('China')
plt.xlabel('Time')
plt.ylabel('Count')
plt.show()

(2) What is the situation in your province or city?

# Cumulative diagnosis in Sichuan
plt.figure()
plt.plot(df.province_confirmedCount)
plt.title('Sichuan')
plt.xlabel('Time')
plt.ylabel('Count')
plt.show()

# Cumulative cure in Sichuan
plt.figure()
plt.plot(df.province_curedCount)
plt.title('China')
plt.xlabel('Time')
plt.ylabel('Count')
plt.show()

# Cumulative Death in Sichuan
plt.figure()
plt.plot(df.province_deadCount)
plt.title('China')
plt.xlabel('Time')
plt.ylabel('Count')
plt.show()

Insert a picture description here

(3) What is the global epidemic situation?

What is the global TOP10 epidemic?
Contrast across continents?
Select a continent of interest to analyze the links, distribution, comparison, and composition of epidemics among countries.

Tip: Pay attention to the use of perspectives, grouping, and integrated knowledge

#  What is the global TOP10 epidemic?

# Maximum number of confirmed cases per country
world_clean
plt.rcParams['font.sans-serif']=['SimHei'] #Used for normal display of Chinese labels
world_clean.sort_values("province_confirmedCount",inplace=False)
world_clean
world_clean = world_clean.drop_duplicates(subset=[ 'province_confirmedCount','countryName'], keep='first', inplace=False)
world_clean
name = world_clean['countryName'].drop_duplicates()
name[:10]
ww = world_clean.copy()
ii = ww.loc[ww['countryName']=='China']
ii.province_confirmedCount
for i in name[:10]:
    ii = ww.loc[ww['countryName']==i]
    plt.figure()
    plt.title(i)
    plt.plot(ii.province_confirmedCount)
    plt.show()

# world_clean['provinceName'].drop_duplicates().values
# world_clean['time'] = world_clean.index
# world_clean = world_clean.drop_duplicates(subset=[ 'updateTime','countryName'], keep='first', inplace=False)
# world_clean = world_clean.drop_duplicates(subset=[ 'province_confirmedCount','countryName'], keep='first', inplace=False)
# world_clean.groupby(by='').max()
# # df['city'].drop_duplicates()
# # list(country_name[-5:])
# # world_clean
# world_clean

# Contrast across continents?
www = world_clean.copy()
name = www['continentName'].drop_duplicates()
name
for i in name:
    ii = www.loc[www['continentName']==i]
    plt.figure()
    plt.title(i)
    plt.plot(ii.province_confirmedCount)
    plt.show()

[External chain picture transfer failed, source station may have anti-theft chain mechanism, it is recommended to save the picture and upload it directly (img-ygkpdj8j-1620734802831)(output_32_0.png)]

[External chain picture transfer failed, source station may have anti-theft chain mechanism, it is recommended to save the picture and upload it directly (img-Ux0n1XLf-1620734802833)(output_32_1.png)]

[External chain picture transfer failed, source station may have anti-theft chain mechanism, it is recommended to save the picture and upload it directly (img-kgAmTK1b-1620734802835)(output_32_2.png)]

[External chain picture transfer failed, source station may have anti-theft chain mechanism, it is recommended to save the picture and upload it directly (img-gL8Hqeku-1620734802839)(output_32_3.png)]

[External chain picture transfer failed, source station may have anti-theft chain mechanism, it is recommended to save the picture and upload it directly (img-ggRosVN2-1620734802843)(output_32_4.png)]

[External chain picture transfer failed, source station may have anti-theft chain mechanism, it is recommended to save the picture and upload it directly (img-GeDEsEFp-1620734802848)(output_32_5.png)]

[External chain picture transfer failed, source station may have anti-theft chain mechanism, it is recommended to save the picture and upload it directly (img-vzStKWuY-1620734802851)(output_32_6.png)]

# Select a continent of interest to analyze the links, distribution, comparison, and composition of epidemics among countries.

ww1 = world_clean.copy()
df_yazhou = ww1.loc[ww1['continentName']=='Asia']
df_yazhou

	continentName	countryName	provinceName	province_confirmedCount	province_suspectedCount	province_curedCount	province_deadCount
updateTime
2021-01-22	Asia	China	China	99667	0.0	92275	4810
2021-01-21	Asia	China	China	99513	0.0	92198	4809
2021-01-20	Asia	China	China	99285	0.0	92130	4808
2021-01-19	Asia	China	China	99094	0.0	92071	4806
2021-01-18	Asia	China	China	98922	0.0	91994	4805
...	...	...	...	...	...	...	...
2020-02-01	Asia	Thailand	Thailand	19	0.0	5	0
2020-02-01	Asia	Malaysia	Malaysia	8	0.0	0	0
2020-02-01	Asia	The United Arab Emirates	The United Arab Emirates	4	0.0	0	0
2020-02-01	Asia	India	India	1	0.0	0	0
2020-01-29	Asia	The United Arab Emirates	The United Arab Emirates	1	0.0	0	0

10960 rows × 7 columns

# There is a lot of duplicate data in a day
# o = o.drop_duplicates(subset=['provinceName', 'updateTime'], keep='first', inplace=False) 
# We only care about the largest number of people diagnosed, sick, treated in each country.
df_yazhou_1 = df_yazhou.drop_duplicates(subset=['countryName'], keep='first', inplace=False) 
df_yazhou_1

	continentName	countryName	provinceName	province_confirmedCount	province_suspectedCount	province_curedCount	province_deadCount
updateTime
2021-01-22	Asia	China	China	99667	0.0	92275	4810
2021-01-22	Asia	Indonesia	Indonesia	965283	0.0	781147	27453
2021-01-22	Asia	Japan	Japan	351020	0.0	279214	4830
2021-01-22	Asia	Pakistan	Pakistan	528891	0.0	482771	11204
2021-01-22	Asia	The Philippines	The Philippines	509887	0.0	467720	10136
2021-01-22	Asia	Cyprus	Cyprus	29472	0.0	2057	176
2021-01-22	Asia	Georgia	Georgia	251071	0.0	238101	2998
2021-01-22	Asia	Jordan	Jordan	318181	0.0	304200	4198
2021-01-22	Asia	Armenia	Armenia	165528	0.0	153857	3021
2021-01-22	Asia	Oman	Oman	132486	0.0	124730	1517
2021-01-22	Asia	Kuwait	Kuwait	159834	0.0	152826	951
2021-01-22	Asia	Qatar	Qatar	148521	0.0	144619	248
2021-01-22	Asia	Thailand	Thailand	13104	0.0	10224	71
2021-01-22	Asia	Maldives	Maldives	14765	0.0	13684	50
2021-01-22	Asia	Cambodia	Cambodia	456	0.0	399	0
2021-01-22	Asia	India	India	10610883	0.0	10162738	152869
2021-01-22	Asia	Iran	Iran	1354520	0.0	1144549	57150
2021-01-22	Asia	Lebanon	Lebanon	269241	0.0	158822	2151
2021-01-22	Asia	Turkey	Turkey	2412505	0.0	2290032	24640
2021-01-22	Asia	Kazakhstan	Kazakhstan	219527	0.0	121347	2956
2021-01-22	Asia	Israel	Israel	582869	0.0	497578	4245
2021-01-22	Asia	The People's Republic of Bangladesh	The People's Republic of Bangladesh	530271	0.0	475074	7966
2021-01-22	Asia	Malaysia	Malaysia	172549	0.0	130152	642
2021-01-22	Asia	The United Arab Emirates	The United Arab Emirates	267258	0.0	239322	766
2021-01-22	Asia	Iraq	Iraq	611407	0.0	576725	12977
2021-01-22	Asia	Myanmar	Myanmar	136166	0.0	119973	3013
2021-01-22	Asia	The Republic of Korea	The Republic of Korea	74262	0.0	61415	1328
2021-01-22	Asia	Sri Lanka	Sri Lanka	56076	0.0	47984	276
2021-01-22	Asia	Azerbaijan	Azerbaijan	228246	0.0	218387	3053
2021-01-22	Asia	Syria	Syria	13313	0.0	6624	858
2021-01-22	Asia	Afghanistan	Afghanistan	54403	0.0	46759	2363
2021-01-22	Asia	Nepal	Nepal	268646	0.0	262868	1979
2021-01-22	Asia	Bahrain	Bahrain	98573	0.0	95240	366
2021-01-22	Asia	Kyrgyzstan	Kyrgyzstan	83585	0.0	79509	1394
2021-01-22	Asia	Saudi Arabia	Saudi Arabia	365775	0.0	357337	6342
2021-01-22	Asia	Republic of Yemen	Republic of Yemen	2119	0.0	350	613
2021-01-22	Asia	Uzbekistan	Uzbekistan	78219	0.0	76624	620
2021-01-22	Asia	Tajikistan	Tajikistan	13714	0.0	12980	91
2021-01-22	Asia	Mongolia	Mongolia	1584	0.0	1046	1
2021-01-22	Asia	Singapore	Singapore	59197	0.0	58926	29
2021-01-22	Asia	Bhutan	Bhutan	851	0.0	660	1
2021-01-22	Asia	Vietnam?	Vietnam?	1546	0.0	1411	35
2021-01-22	Asia	Timor-Leste	Timor-Leste	53	0.0	30	0
2021-01-22	Asia	Palestine	Palestine	65802	0.0	32944	561
2021-01-22	Asia	Brunei	Brunei	174	0.0	169	3
2021-01-22	Asia	Laos	Laos	41	0.0	41	0
2020-05-03	Asia	Turkmenistan	Turkmenistan	0	0.0	0	0
2020-03-15	Asia	Reunion	Reunion	3	0.0	0	0

yazhou1 = df_yazhou_1[['countryName','province_confirmedCount']]
yazhou1.set_index('countryName',inplace=True)
yazhou1
list(yazhou1.index)

['China',
 'Indonesia',
 'Japan',
 'Pakistan',
 'The Philippines',
 'Cyprus',
 'Georgia',
 'Jordan',
 'Armenia',
 'Oman',
 'Kuwait',
 'Qatar',
 'Thailand',
 'Maldives',
 'Cambodia',
 'India',
 'Iran',
 'Lebanon',
 'Turkey',
 'Kazakhstan',
 'Israel',
 'The People's Republic of Bangladesh',
 'Malaysia',
 'The United Arab Emirates',
 'Iraq',
 'Myanmar',
 'The Republic of Korea',
 'Sri Lanka',
 'Azerbaijan',
 'Syria',
 'Afghanistan',
 'Nepal',
 'Bahrain',
 'Kyrgyzstan',
 'Saudi Arabia',
 'Republic of Yemen',
 'Uzbekistan',
 'Tajikistan',
 'Mongolia',
 'Singapore',
 'Bhutan',
 'Vietnam?',
 'Timor-Leste',
 'Palestine',
 'Brunei',
 'Laos',
 'Turkmenistan',
 'Reunion']

# province_confirmedCount

plt.figure(figsize=(6,9)) #Adjust graphic size
plt.pie(yazhou1)
plt.legend(list(yazhou1.index))
plt.show()


# province_curedCount
yazhou2 = df_yazhou_1[['countryName','province_curedCount']]
yazhou2.set_index('countryName',inplace=True)
yazhou2
list(yazhou2.index)
plt.figure(figsize=(6,9)) #Adjust graphic size
plt.pie(yazhou2)
plt.legend(list(yazhou2.index))
plt.show()


# province_deadCount
yazhou3 = df_yazhou_1[['countryName','province_deadCount']]
yazhou3.set_index('countryName',inplace=True)
yazhou3
list(yazhou3.index)
plt.figure(figsize=(6,9)) #Adjust graphic size
plt.pie(yazhou3)
plt.legend(list(yazhou3.index))
plt.show()

(4) Give your judgment on the epidemic trend in the next half year based on your analysis results, and what suggestions do you have for individuals and society in fighting the epidemic?

from pyecharts import Map,Geo
import pandas as pd
 
# df = pd.read_csv("Major Cities Annual Data. csv",encoding="gb2312")
city = china['provinceName']
count = china['province_confirmedCount']
hospital_map = Geo("Number of urban areas","2021-01-22")#Header and subtitle hospital_map.use_theme=("white")
hospital_map.add("",city,count,maptype="china",is_visualmap=True,is_lable_show=True,visual_range=[min(count),max(count)],visual_text_color="#000")
hospital_map.render("hospital map.html")

Additional analysis (optional, but better results)

Additional analysis can be done using unlimited libraries, such as seaborn, pyecharts, and so on.

Children's shoes, free play!! (There's really no idea, so go to Baidu epidemic map or other sites for epidemic analysis to get inspiration.)

For example, such drops...

[External chain picture transfer failed, source station may have anti-theft chain mechanism, it is recommended to save the picture and upload it directly (img-a711p47D-1620734802865)(images/chinamap.png)]

Or, drop like this...

[External chain picture transfer failed, source station may have anti-theft chain mechanism, it is recommended to save the picture and upload it directly (img-BkjHqX7o-1620734802866)(images/worldmap.png)]

Or, drop like this...

[External chain picture transfer failed, source station may have anti-theft chain mechanism, it is recommended to save the picture and upload it directly (img-qwYlGObf-1620734802867)(images/meiguitu.png)]

Major man, wait for you to fight!

Tags: Big Data Python Data Analysis Visualization data visualization

Posted by Afrojojo on Thu, 02 Jun 2022 20:19:38 +0530

Programs Team

[Python] Using Python for data analysis--a case study of a new coronavirus epidemic

[Python] Using Python for data analysis--a case study of a new coronavirus epidemic

Important Notes

1. Ask questions

2. Understanding data

3. Data cleaning

(1) Basic data processing

Selection of domestic epidemic data (the final selected data is named china)

Selection of epidemic data in your province or city (the final selected data is named myhome)

Global epidemic data selection (the final selected data is named world)

4. Data analysis and visualization

Basic analysis

(1) What is the trend of cumulative diagnosis/cure/death over time throughout the country?

(2) What is the situation in your province or city?

(3) What is the global epidemic situation?

(4) Give your judgment on the epidemic trend in the next half year based on your analysis results, and what suggestions do you have for individuals and society in fighting the epidemic?

Additional analysis (optional, but better results)

Hot Tags