mirror of
https://github.com/LCTT/TranslateProject.git
synced 2025-02-25 00:50:15 +08:00
Article has been translated
Article has been translated
This commit is contained in:
zxp
parent
d8673b7c66
commit
ddff56ea75
@ -1,174 +0,0 @@
|
||||
[#]: collector: (lujun9972)
|
||||
[#]: translator: (zhangxiangping)
|
||||
[#]: reviewer: ( )
|
||||
[#]: publisher: ( )
|
||||
[#]: url: ( )
|
||||
[#]: subject: (How I use Python to map the global spread of COVID-19)
|
||||
[#]: via: (https://opensource.com/article/20/4/python-map-covid-19)
|
||||
[#]: author: (AnuragGupta https://opensource.com/users/999anuraggupta)
|
||||
|
||||
我如何使用Python绘制COVID-19的全球扩散图
|
||||
======
|
||||
使用这些开源框架创建一个彩色地图，显示病毒的可能的传播路径。![Globe up in the clouds][1]
|
||||
|
||||
|
||||
|
||||
Create a color coded geographic map of the potential spread of the virus
|
||||
using these open source scripts.
|
||||
![Globe up in the clouds][1]
|
||||
|
||||
The spread of disease is a real concern for a world in which global travel is commonplace. A few organizations track significant epidemics (and any pandemic), and fortunately, they publish their work as open data. The raw data can be difficult for humans to process, though, and that's why data science is so vital. For instance, it could be useful to visualize the worldwide spread of COVID-19 with Python and Pandas.
|
||||
|
||||
It can be hard to know where to start when you're faced with large amounts of raw data. The more you do it, however, the more patterns begin to emerge. Here's a common scenario, applied to COVID-19 data:
|
||||
|
||||
1. Download COVID-19 country spread daily data into a Pandas DataFrame object from GitHub. For this, you need the Python Pandas library.
|
||||
2. Process and clean the downloaded data and make it suitable for visualizing. The downloaded data (as you will see for yourself) is in quite good condition. The one problem with this data is that it uses the names of countries, but it's better to use three-digit ISO 3 codes. To generate the three-digit ISO 3 codes, use a small Python library called pycountry. Having generated these codes, you can add an extra column to our DataFrame and populate it with these codes.
|
||||
3. Finally, for the visualization, use the **express** module of a library called Plotly. This article uses what are called choropleth maps (available in Plotly) to visualize the worldwide spread of the disease.
|
||||
|
||||
|
||||
|
||||
### Step 1: Corona data
|
||||
|
||||
We will download the latest corona data from:
|
||||
|
||||
<https://raw.githubusercontent.com/datasets/covid-19/master/data/countries-aggregated.csv>
|
||||
|
||||
We will load the data directly into a Pandas DataFrame. Pandas provides a function, **read_csv()**, which can take a URL and return a DataFrame object as shown below:
|
||||
|
||||
|
||||
```
|
||||
import pycountry
|
||||
import plotly.express as px
|
||||
import pandas as pd
|
||||
URL_DATASET = r'<https://raw.githubusercontent.com/datasets/covid-19/master/data/countries-aggregated.csv>'
|
||||
df1 = pd.read_csv(URL_DATASET)
|
||||
print(df1.head(3)) # Get first 3 entries in the dataframe
|
||||
print(df1.tail(3)) # Get last 3 entries in the dataframe
|
||||
```
|
||||
|
||||
The screenshot of output (on Jupyter) is:
|
||||
|
||||
![Jupyter screenshot][2]
|
||||
|
||||
From output, you can see that the DataFrame (df1) has the following columns:
|
||||
|
||||
1. Date
|
||||
2. Country
|
||||
3. Confirmed
|
||||
4. Recovered
|
||||
5. Dead
|
||||
|
||||
|
||||
|
||||
Further, you can see that the **Date** column has entries starting from January 22 to March 31. This database is updated daily, so you will get the current values.
|
||||
|
||||
### Step 2: Cleaning and modifying the data frame
|
||||
|
||||
We need to add another column to this DataFrame, which has the three-letter ISO alpha-3 codes. To do this, I followed these steps:
|
||||
|
||||
1. Create a list of all countries in the database. This was required because in the **df**, in the column **Country**, each country was figuring for each date. So in effect, the **Country** column had multiple entries for each country. To do this, I used the **unique().tolist()** functions.
|
||||
2. Then I took a dictionary **d_country_code** (initially empty) and populated it with keys consisting of country names and values consisting of their three-letter ISO codes.
|
||||
3. To generate the three-letter ISO code for a country, I used the function **pycountry.countries.search_fuzzy(country)**. You need to understand that the return value of this function is a "list of **Country** objects." I passed the return value of this function to a name country_data. Further, in this list of objects, the first object i.e., at index 0, is the best fit. Further, this **\** object has an attribute **alpha_3**. So, I can "access" the 3 letter ISO code by using **country_data[0].alpha_3**. However, it is possible that some country names in the DataFrame may not have a corresponding ISO code (For example, disputed territories). So, for such countries, I gave an ISO code of "i.e. a blank string. Further, you need to wrap this code in a try-except block. The statement: **print(_‘could not add ISO 3 code for ->'_, country)** will give a printout of those countries for which the ISO 3 codes could not be found. In fact, you will find such countries as shown with white color in the final output.
|
||||
4. Having got the three-letter ISO code for each country (or an empty string for some), I added the country name (as key) and its corresponding ISO code (as value) to the dictionary **d_country_code**. For adding these, I used the **update()** method of the Python dictionary object.
|
||||
5. Having created a dictionary of country names and their codes, I added them to the DataFrame using a simple for loop.
|
||||
|
||||
|
||||
|
||||
### Step 3: Visualizing the spread using Plotly
|
||||
|
||||
A choropleth map is a map composed of colored polygons. It is used to represent spatial variations of a quantity. We will use the express module of Plotly conventionally called **px**. Here we show you how to create a choropleth map using the function: **px.choropleth**.
|
||||
|
||||
The signature of this function is:
|
||||
|
||||
|
||||
```
|
||||
`plotly.express.choropleth(data_frame=None, lat=None, lon=None, locations=None, locationmode=None, geojson=None, featureidkey=None, color=None, hover_name=None, hover_data=None, custom_data=None, animation_frame=None, animation_group=None, category_orders={}, labels={}, color_discrete_sequence=None, color_discrete_map={}, color_continuous_scale=None, range_color=None, color_continuous_midpoint=None, projection=None, scope=None, center=None, title=None, template=None, width=None, height=None)`
|
||||
```
|
||||
|
||||
The noteworthy points are that the **choropleth()** function needs the following things:
|
||||
|
||||
1. A geometry in the form of a **geojson** object. This is where things are a bit confusing and not clearly mentioned in its documentation. You may or may not provide a **geojson** object. If you provide a **geojson** object, then that object will be used to plot the earth features, but if you don't provide a **geojson** object, then the function will, by default, use one of the built-in geometries. (In our example here, we will use a built-in geometry, so we won't provide any value for the **geojson** argument)
|
||||
2. A pandas DataFrame object for the attribute **data_frame**. Here we provide our DataFrame ie **df1** we created earlier.
|
||||
3. We will use the data of **Confirmed** column to decide the color of each country polygon.
|
||||
4. Further, we will use the **Date** column to create the **animation_frame**. Thus as we slide across the dates, the colors of the countries will change as per the values in the **Confirmed** column.
|
||||
|
||||
|
||||
|
||||
The complete code is given below:
|
||||
|
||||
|
||||
```
|
||||
import pycountry
|
||||
import plotly.express as px
|
||||
import pandas as pd
|
||||
# ----------- Step 1 ------------
|
||||
URL_DATASET = r'<https://raw.githubusercontent.com/datasets/covid-19/master/data/countries-aggregated.csv>'
|
||||
df1 = pd.read_csv(URL_DATASET)
|
||||
# print(df1.head) # Uncomment to see what the dataframe is like
|
||||
# ----------- Step 2 ------------
|
||||
list_countries = df1['Country'].unique().tolist()
|
||||
# print(list_countries) # Uncomment to see list of countries
|
||||
d_country_code = {} # To hold the country names and their ISO
|
||||
for country in list_countries:
|
||||
try:
|
||||
country_data = pycountry.countries.search_fuzzy(country)
|
||||
# country_data is a list of objects of class pycountry.db.Country
|
||||
# The first item ie at index 0 of list is best fit
|
||||
# object of class Country have an alpha_3 attribute
|
||||
country_code = country_data[0].alpha_3
|
||||
d_country_code.update({country: country_code})
|
||||
except:
|
||||
print('could not add ISO 3 code for ->', country)
|
||||
# If could not find country, make ISO code ' '
|
||||
d_country_code.update({country: ' '})
|
||||
|
||||
# print(d_country_code) # Uncomment to check dictionary
|
||||
|
||||
# create a new column iso_alpha in the df
|
||||
# and fill it with appropriate iso 3 code
|
||||
for k, v in d_country_code.items():
|
||||
df1.loc[(df1.Country == k), 'iso_alpha'] = v
|
||||
|
||||
# print(df1.head) # Uncomment to confirm that ISO codes added
|
||||
# ----------- Step 3 ------------
|
||||
fig = px.choropleth(data_frame = df1,
|
||||
locations= "iso_alpha",
|
||||
color= "Confirmed", # value in column 'Confirmed' determines color
|
||||
hover_name= "Country",
|
||||
color_continuous_scale= 'RdYlGn', # color scale red, yellow green
|
||||
animation_frame= "Date")
|
||||
|
||||
fig.show()
|
||||
```
|
||||
|
||||
The output is something like the following:
|
||||
|
||||
![Map][3]
|
||||
|
||||
You can download and run the [complete code][4].
|
||||
|
||||
To wrap up, here are some excellent resources on choropleth in Plotly:
|
||||
|
||||
* <https://github.com/plotly/plotly.py/blob/master/doc/python/choropleth-maps.md>
|
||||
* [https://plotly.com/python/reference/#choropleth][5]
|
||||
|
||||
|
||||
|
||||
--------------------------------------------------------------------------------
|
||||
|
||||
via: https://opensource.com/article/20/4/python-map-covid-19
|
||||
|
||||
作者：[AnuragGupta][a]
|
||||
选题：[lujun9972][b]
|
||||
译者：[zhangxiangping](https://github.com/zxp93)
|
||||
校对：[校对者ID](https://github.com/校对者ID)
|
||||
|
||||
本文由 [LCTT](https://github.com/LCTT/TranslateProject) 原创编译，[Linux中国](https://linux.cn/) 荣誉推出
|
||||
|
||||
[a]: https://opensource.com/users/999anuraggupta
|
||||
[b]: https://github.com/lujun9972
|
||||
[1]: https://opensource.com/sites/default/files/styles/image-full-size/public/lead-images/cloud-globe.png?itok=_drXt4Tn (Globe up in the clouds)
|
||||
[2]: https://opensource.com/sites/default/files/uploads/jupyter_screenshot.png (Jupyter screenshot)
|
||||
[3]: https://opensource.com/sites/default/files/uploads/map_2.png (Map)
|
||||
[4]: https://github.com/ag999git/jupyter_notebooks/blob/master/corona_spread_visualization
|
||||
[5]: tmp.azs72dmHFd#choropleth
|
||||
@ -0,0 +1,154 @@
|
||||
[#]: collector: (lujun9972)
|
||||
[#]: translator: (zhangxiangping)
|
||||
[#]: reviewer: ( )
|
||||
[#]: publisher: ( )
|
||||
[#]: url: ( )
|
||||
[#]: subject: (How I use Python to map the global spread of COVID-19)
|
||||
[#]: via: (https://opensource.com/article/20/4/python-map-covid-19)
|
||||
[#]: author: (AnuragGupta https://opensource.com/users/999anuraggupta)
|
||||
|
||||
如何使用Python绘制COVID-19的全球扩散图
|
||||
======
|
||||
使用这些开源框架创建一个彩色地图，显示病毒的可能的传播路径。![Globe up in the clouds][1]
|
||||
|
||||
在全球范围内的旅行中，疾病的传播是一个令人担忧的问题。一些组织会跟踪重大的流行病（还有所有普遍的流行病），并将他们的跟踪工作获得的数据公开出来。这些原生的数据对人来说可能很难处理，这就是为什么数据科学如此重要的原因。比如，将COVID-19在全球范围内的传播路径用Python和Pandas进行可视化可能对这些数据的分析有所帮助。
|
||||
|
||||
最开始，当面对如此大数量的原始数据时可能难以下手。但当你开始处理数据之后，慢慢地就会发现一些处理数据的方式。下面是用于处理COVID-19数据的一些常见的情况：
|
||||
1. 从Github上下载COVID-19的国际间每日传播数据，保存为一个Pandas中的DataFrame对象。这时你需要使用Python中的Pandas库。
|
||||
2. 处理并清理下载好的数据，使其满足可视化数据的输入格式。所下载的数据的情况很好（数据规整）。这个数据有一个问题是它用国家的名字来标识国家，但最好是使用ISO 3码（国家代码表）来标识国家。为了生成三位ISO 3码，可是使用pycountry这个Python库。生成了这些代码之后，可以在原有的DataFrame上增加一列，然后用这些代码填充进去。
|
||||
3. 最后为了可视化，使用Plotly库中的**express**模块。这篇文章是使用choropleth maps（在Plotly库中）来对疾病在全球的传播进行可视化的。
|
||||
|
||||
### 第一步：Corona数据
|
||||
|
||||
从下面这个网站上下载最新的corona数据（译者注：2020-12-14仍可访问，有墙）：
|
||||
|
||||
<https://raw.githubusercontent.com/datasets/covid-19/master/data/countries-aggregated.csv>
|
||||
|
||||
我们之间将这个下载好的数据载入为Pandas的DataFrame。Pandas提供了一个函数， **read_csv()**，可以直接使用URL读取数据，并返回一个DataFrame对象，具体如下所示：
|
||||
|
||||
```
|
||||
import pycountry
|
||||
import plotly.express as px
|
||||
import pandas as pd
|
||||
URL_DATASET = r'<https://raw.githubusercontent.com/datasets/covid-19/master/data/countries-aggregated.csv>'
|
||||
df1 = pd.read_csv(URL_DATASET)
|
||||
print(df1.head(3)) # Get first 3 entries in the dataframe
|
||||
print(df1.tail(3)) # Get last 3 entries in the dataframe
|
||||
```
|
||||
在Jupyter上的输出截图：
|
||||
|
||||
![Jupyter screenshot][2]
|
||||
|
||||
从这个输出可以看到这个DataFrame（df1）包括以下几列数据：
|
||||
1. Date
|
||||
2. Country
|
||||
3. Confirmed
|
||||
4. Recovered
|
||||
5. Dead
|
||||
|
||||
之后还可以看到**Date**这一列包含了从1月22日到3月31日的条目信息。这个数据是每天更新的，所以你会得到你当天的值。
|
||||
|
||||
### 第二步：清理和修改数据框
|
||||
|
||||
我们要往这个DataFrame中增加一列数据，就是那个包含了ISO 3编码。可以通过以下三步完成这个任务：
|
||||
|
||||
1. 创建一个包含所有国家的list。因为在**df**的**Country**列中，国家都是每个日期就重复一次。所以实际上**Country**列中对每个国家就会有多个条目。我使用**unique().tolist()**函数完成这个任务。
|
||||
2. 我使用**d_country_code**字典对象（初始为空），然后将其键设置为国家的名称，然后它的值设置为其对应的ISO 3编码。
|
||||
3. 我使用**pycountry.countries.search_fuzzy(country)**为每个国家生成ISO 3编码。你需要明白的是这个函数的返回值是一个**Country**对象的list。我将这个函数的返回值赋给country_data对象。之以这个对象的第一个元素（序号0）为例。这个**\**对象有一个**alpha_3**属性。所以我使用**country_data[0].alpha_3**就能"获得"第一个元素的ISO 3编码。然而，在这个DataFrame中有些国家的名称可能没有对应的ISO 3编码（比如有争议的领土）。那么对这些“国家”，我就用一个空白字符串来替代ISO编码。你也可以用一个try-except代码来替换这部分。except中的语句可以写：**print(_‘could not add ISO 3 code for ->'_, country)**。这样就能在找不到这些“国家”对应的ISO 3编码时给出一个输出提示。实际上，你会发现这些“国家”会在最后的输出中用白色来表示。
|
||||
4. 在获得了每个国家的ISO 3编码（有些是空白字符串）之后，我把这些国家的名称（作为键）还有国家对应的ISO 3编码（作为值）添加到之前的字典**d_country_code**中。可以使用Python中字典对象的**update()**方法来完成这个任务。
|
||||
5. 在创建好了一个包含国家名称和对应ISO 3编码的字典之后，我使用一个简单的循环将他们加入到DataFrame中。
|
||||
|
||||
|
||||
### 第三步：使用Plotly可视化传播路径
|
||||
|
||||
choropleth图是一个由彩色多边形组成的图。它常常用来表示一个变量在空间中的变化。我们使用Plotly中的**px**模块来创建choropleth图，具体函数为：**px.choropleth**。
|
||||
|
||||
这个函数的所包含的参数如下：
|
||||
|
||||
```
|
||||
`plotly.express.choropleth(data_frame=None, lat=None, lon=None, locations=None, locationmode=None, geojson=None, featureidkey=None, color=None, hover_name=None, hover_data=None, custom_data=None, animation_frame=None, animation_group=None, category_orders={}, labels={}, color_discrete_sequence=None, color_discrete_map={}, color_continuous_scale=None, range_color=None, color_continuous_midpoint=None, projection=None, scope=None, center=None, title=None, template=None, width=None, height=None)`
|
||||
```
|
||||
|
||||
**choropleth()**这个函数还有几点需要注意：
|
||||
|
||||
1. **geojson**是一个geometry对象（上面函数第六个参数）。这个对象有点让人困扰，因为在函数文档中没有明确地提到这个对象。你可以提供，也可以不提供**geojson**对象。如果你提供了**geojson**对象，那么这个对象就会被用来绘制地球特征，如果不提供**geojson**对象，那这个函数默认就会使用一个内建的geometry对象。（在我们的实验中，我们使用内建的geometry对象，因此我们不会为**geojson**参数提供值）
|
||||
2. DataFrame对象有一个**data_frame**属性，在这里我们先前就提供了一个我们创建好的**df1**。
|
||||
3. 我们用**Confirmed**（确诊）来决定每个国家多边形的颜色。
|
||||
4. 最后，我们**Date**列创建一个**animation_frame**。这样我们就能通过日期来划分数据，国家的颜色会随着**Confirmed**的变化而变化。
|
||||
|
||||
最后完整的代码如下：
|
||||
|
||||
```
|
||||
import pycountry
|
||||
import plotly.express as px
|
||||
import pandas as pd
|
||||
# ----------- Step 1 ------------
|
||||
URL_DATASET = r'<https://raw.githubusercontent.com/datasets/covid-19/master/data/countries-aggregated.csv>'
|
||||
df1 = pd.read_csv(URL_DATASET)
|
||||
# print(df1.head) # Uncomment to see what the dataframe is like
|
||||
# ----------- Step 2 ------------
|
||||
list_countries = df1['Country'].unique().tolist()
|
||||
# print(list_countries) # Uncomment to see list of countries
|
||||
d_country_code = {} # To hold the country names and their ISO
|
||||
for country in list_countries:
|
||||
try:
|
||||
country_data = pycountry.countries.search_fuzzy(country)
|
||||
# country_data is a list of objects of class pycountry.db.Country
|
||||
# The first item ie at index 0 of list is best fit
|
||||
# object of class Country have an alpha_3 attribute
|
||||
country_code = country_data[0].alpha_3
|
||||
d_country_code.update({country: country_code})
|
||||
except:
|
||||
print('could not add ISO 3 code for ->', country)
|
||||
# If could not find country, make ISO code ' '
|
||||
d_country_code.update({country: ' '})
|
||||
|
||||
# print(d_country_code) # Uncomment to check dictionary
|
||||
|
||||
# create a new column iso_alpha in the df
|
||||
# and fill it with appropriate iso 3 code
|
||||
for k, v in d_country_code.items():
|
||||
df1.loc[(df1.Country == k), 'iso_alpha'] = v
|
||||
|
||||
# print(df1.head) # Uncomment to confirm that ISO codes added
|
||||
# ----------- Step 3 ------------
|
||||
fig = px.choropleth(data_frame = df1,
|
||||
locations= "iso_alpha",
|
||||
color= "Confirmed", # value in column 'Confirmed' determines color
|
||||
hover_name= "Country",
|
||||
color_continuous_scale= 'RdYlGn', # color scale red, yellow green
|
||||
animation_frame= "Date")
|
||||
|
||||
fig.show()
|
||||
```
|
||||
这段代码的输出就是下面这个图的内容：
|
||||
|
||||
![Map][3]
|
||||
|
||||
你可以从这里下载并运行完整代码[complete code][4]。
|
||||
|
||||
最后，这里还有一些关于Plotly绘制choropleth图的不错的资源。
|
||||
|
||||
* <https://github.com/plotly/plotly.py/blob/master/doc/python/choropleth-maps.md>
|
||||
* [https://plotly.com/python/reference/#choropleth][5]
|
||||
|
||||
|
||||
|
||||
--------------------------------------------------------------------------------
|
||||
|
||||
via: https://opensource.com/article/20/4/python-map-covid-19
|
||||
|
||||
作者：[AnuragGupta][a]
|
||||
选题：[lujun9972][b]
|
||||
译者：[zhangxiangping](https://github.com/zxp93)
|
||||
校对：[校对者ID](https://github.com/校对者ID)
|
||||
|
||||
本文由 [LCTT](https://github.com/LCTT/TranslateProject) 原创编译，[Linux中国](https://linux.cn/) 荣誉推出
|
||||
|
||||
[a]: https://opensource.com/users/999anuraggupta
|
||||
[b]: https://github.com/lujun9972
|
||||
[1]: https://opensource.com/sites/default/files/styles/image-full-size/public/lead-images/cloud-globe.png?itok=_drXt4Tn (Globe up in the clouds)
|
||||
[2]: https://opensource.com/sites/default/files/uploads/jupyter_screenshot.png (Jupyter screenshot)
|
||||
[3]: https://opensource.com/sites/default/files/uploads/map_2.png (Map)
|
||||
[4]: https://github.com/ag999git/jupyter_notebooks/blob/master/corona_spread_visualization
|
||||
[5]: tmp.azs72dmHFd#choropleth
|
||||
Loading…
Reference in New Issue
Block a user