Typhoons are raging recently. Let us use Python to obtain weather data and analyze where Typhoon Kanu is going.

The typhoon has been raging recently and has entered our country's 24-hour warning line! Where is Typhoon Kanu going?

As a Python programmer, although I can't help, I can still pay attention to it from time to time. By the way, I pray that the typhoon blows to a small island in the east where I have a good life.

So I spent a minute writing a code to obtain weather data in Python, and then analyzed the data to see if it was going to work.

First we need to prepare these

Software Environment

• python
• pycharm

module

• requests # Send requests
• parsel # Parse data
• The complete source code and video explanation can be obtained directly from the business card at the end of the article.

These are third-party modules and need to be installed manually. If not installed, install them with pip.

Knowledge points and process ideas

Knowledge points

• Dynamic data packet capture
• requestsSend a request
• Structured + unstructured data parsing

crawler process

Introduction:
Simulate as a browser (client) and send network requests to the 2345 server.
Function:
Batch data collection/simulate user behavior

Case implementation

1. Idea analysis

Find data sources
Static data
Dynamic data: Shortcuts

2. Code implementation

1. Access data source address through code
2. After accessing, get the data content
3. Take out the content we need from the data content and eliminate the unnecessary content.
4. Save to form

Code display

Weather data acquisition

import requests     # 发送请求的第三方库 用来访问网站的
import parsel       # 第三方库 提取数据的
import csv          # 内置模块 无需安装
 
f = open('tianqi.csv', mode='a', encoding='utf-8', newline='')
csv_writer = csv.writer(f)
 
for year in range(2013, 2023):
    for month in range(1, 13):
        url = f'https://***.com/Pc/GetHistory?areaInfo%5BareaId%5D=54511&areaInfo%5BareaType%5D=2&date%5Byear%5D={
      
      year}&date%5Bmonth%5D={
      
      month}'
        # 1. 通过代码的方式访问 数据来源地址
        response = requests.get(url)
        # 2. 访问之后 将 数据内容 拿到
        json_data = response.json()
        # 3. 将数据内容中 我们需要的内容取出来 不需要的内容 就剔除掉
        html_data = json_data['data']
        select = parsel.Selector(html_data)
        trs = select.css('tr')  
        for tr in trs[1:]:
            tds = tr.css('td::text').getall()
            # 4. 保存到表格当中
            csv_writer.writerow(tds)

Data analysis part

Import package

import pandas as pd
import datetime
from pyecharts import options as opts
from pyecharts.charts import *
from pyecharts.commons.utils import JsCode

Read in data

data = pd.read_csv('天气.csv')
data

Data preview

data.sample(5)
data.info()

Split date/week

data[['日期','星期']] = data['日期'].str.split(' ',expand=True,n=1)
data

Remove extra characters

data[['最高温度','最低温度']] = data[['最高温度','最低温度']].apply(lambda x: x.str.replace('°','').replace('', '0'))
data.head()

Calculate snowy weather

data.loc[data['天气'].str.contains('雪'),'下雪吗']='是'
data.fillna('否',inplace=True)

Split date time

data['日期'] = pd.to_datetime(data['日期'])
data[['最高温度','最低温度']] = data[['最高温度','最低温度']].astype('int')

data['年份'] = data['日期'].dt.year
data['月份'] = data['日期'].dt.month
data['日'] = data['日期'].dt.day
# 预览
data.sample(5)

Time for first snowfall in each city

s_data = data[data['下雪吗']=='是']
s_data[(s_data['月份']>=9)].groupby('年份').first().reset_index()

Distribution of snowy weather in various cities

s_data.groupby(['城市','年份'])['日期'].count().to_frame('下雪天数').reset_index()

Make a pivot table

data_bj = data[(data['年份'] == 2021) & (data['城市'] == '北京')]
data_bj = data_bj.groupby(['月份','天气'], as_index=False)['日期'].count()

data_pivot =  pd.pivot(data_bj,
                values='日期',
                index='月份',
                columns='天气')
data_pivot = data_pivot.astype('float')
# 按照 索引年月倒序排序
data_pivot.sort_index(ascending=False,inplace=True)

data_pivot

Distribution of weather heat maps in Beijing, Shanghai, Guangzhou and Shenzhen in October 2021

import matplotlib.pyplot as plt
import matplotlib.colors as mcolors
import seaborn as sns

#设置全局默认字体 为 雅黑
plt.rcParams['font.family'] = ['Microsoft YaHei'] 
# 设置全局轴标签字典大小
plt.rcParams["axes.labelsize"] = 14  
# 设置背景
sns.set_style("darkgrid",{
    
    "font.family":['Microsoft YaHei', 'SimHei']})  
# 设置画布长宽 和 dpi
plt.figure(figsize=(18,8),dpi=100)
# 自定义色卡
cmap = mcolors.LinearSegmentedColormap.from_list("n",['#95B359','#D3CF63','#E0991D','#D96161','#A257D0','#7B1216']) 
# 绘制热力图

ax = sns.heatmap(data_pivot, cmap=cmap, vmax=30, 
                 annot=True, # 热力图上显示数值
                 linewidths=0.5,
                ) 
# 将x轴刻度放在最上面
ax.xaxis.set_ticks_position('top') 
plt.title('北京最近10个月天气分布',fontsize=16) #图片标题文本和字体大小
plt.show()

data_gz= data[(data['年份'] == 2021) & (data['城市'] == '广州')]
data_gz = data_gz.groupby(['月份','天气'], as_index=False)['日期'].count()
data_sz= data[(data['年份'] == 2021) & (data['城市'] == '深圳')]
data_sz = data_sz.groupby(['月份','天气'], as_index=False)['日期'].count()
data_sh= data[(data['年份'] == 2021) & (data['城市'] == '上海')]
data_sh = data_sh.groupby(['月份','天气'], as_index=False)['日期'].count()

data_pivot_sz =  pd.pivot(data_sz,
                values='日期',
                index='月份',
                columns='天气')
data_pivot_sz = data_pivot_sz.astype('float')
# 按照 索引年月倒序排序
data_pivot_sz.sort_index(ascending=False,inplace=True)

#设置全局默认字体 为 雅黑
plt.rcParams['font.family'] = ['Microsoft YaHei'] 
# 设置全局轴标签字典大小
plt.rcParams["axes.labelsize"] = 14  
# 设置背景
sns.set_style("darkgrid",{
    
    "font.family":['Microsoft YaHei', 'SimHei']})  
# 设置画布长宽 和 dpi
plt.figure(figsize=(18,8),dpi=100)
# 自定义色卡
cmap = mcolors.LinearSegmentedColormap.from_list("n",['#95B359','#D3CF63','#E0991D','#D96161','#A257D0','#7B1216']) 
# 绘制热力图

ax_sz = sns.heatmap(data_pivot_sz, cmap=cmap, vmax=31, 
                 annot=True, # 热力图上显示数值
                 linewidths=0.5,
                ) 
# 将x轴刻度放在最上面
ax_sz.xaxis.set_ticks_position('top') 
plt.title('深圳最近10个月天气分布',fontsize=16) #图片标题文本和字体大小
plt.show()

data_pivot_gz =  pd.pivot(data_gz,
                values='日期',
                index='月份',
                columns='天气')
data_pivot_gz = data_pivot_gz.astype('float')
# 按照 索引年月倒序排序
data_pivot_gz.sort_index(ascending=False,inplace=True)

#设置全局默认字体 为 雅黑
plt.rcParams['font.family'] = ['Microsoft YaHei'] 
# 设置全局轴标签字典大小
plt.rcParams["axes.labelsize"] = 14  
# 设置背景
sns.set_style("darkgrid",{
    
    "font.family":['Microsoft YaHei', 'SimHei']})  
# 设置画布长宽 和 dpi
plt.figure(figsize=(18,8),dpi=100)
# 自定义色卡
cmap = mcolors.LinearSegmentedColormap.from_list("n",['#95B359','#D3CF63','#E0991D','#D96161','#A257D0','#7B1216']) 
# 绘制热力图

ax_sz = sns.heatmap(data_pivot_gz, cmap=cmap, vmax=31, 
                 annot=True, # 热力图上显示数值
                 linewidths=0.5,
                ) 
# 将x轴刻度放在最上面
ax_sz.xaxis.set_ticks_position('top') 
plt.title('广州最近10个月天气分布',fontsize=16) #图片标题文本和字体大小
plt.show()

data_pivot_sh =  pd.pivot(data_sh,
                values='日期',
                index='月份',
                columns='天气')
data_pivot_sh = data_pivot_sh.astype('float')
# 按照 索引年月倒序排序
data_pivot_sh.sort_index(ascending=False,inplace=True)

#设置全局默认字体 为 雅黑
plt.rcParams['font.family'] = ['Microsoft YaHei'] 
# 设置全局轴标签字典大小
plt.rcParams["axes.labelsize"] = 14  
# 设置背景
sns.set_style("darkgrid",{
    
    "font.family":['Microsoft YaHei', 'SimHei']})  
# 设置画布长宽 和 dpi
plt.figure(figsize=(18,8),dpi=100)
# 自定义色卡
cmap = mcolors.LinearSegmentedColormap.from_list("n",['#95B359','#D3CF63','#E0991D','#D96161','#A257D0','#7B1216']) 
# 绘制热力图

ax_sz = sns.heatmap(data_pivot_sh, cmap=cmap, vmax=31, 
                 annot=True, # 热力图上显示数值
                 linewidths=0.5,
                ) 
# 将x轴刻度放在最上面
ax_sz.xaxis.set_ticks_position('top') 
plt.title('上海最近10个月天气分布',fontsize=16) #图片标题文本和字体大小
plt.show()

data_bj = data[(data['城市']=='北京') & (data['年份'] == 2021)]
data_bj['日期'] = pd.to_datetime(data_bj.日期,format="%Y年%m月%d日")
data_bj = data_bj.sort_values(by='日期',ascending=True)

Daily maximum and minimum temperature changes in Beijing in 2021

color0 = ['#FF76A2','#24ACE6']
color_js0 = """new echarts.graphic.LinearGradient(0, 1, 0, 0,
    [{offset: 0, color: '#FFC0CB'}, {offset: 1, color: '#ed1941'}], false)"""
color_js1 = """new echarts.graphic.LinearGradient(0, 1, 0, 0,
    [{offset: 0, color: '#FFFFFF'}, {offset: 1, color: '#009ad6'}], false)"""

tl = Timeline()
for i in range(0,len(data_bj)):
    coordy_high = list(data_bj['最高温度'])[i]
    coordx = list(data_bj['日期'])[i]
    coordy_low = list(data_bj['最低温度'])[i]
    x_max = list(data_bj['日期'])[i]+datetime.timedelta(days=10)
    y_max = int(max(list(data_bj['最高温度'])[0:i+1]))+3
    y_min = int(min(list(data_bj['最低温度'])[0:i+1]))-3
    title_date = list(data_bj['日期'])[i].strftime('%Y-%m-%d')
    c = (
        Line(
            init_opts=opts.InitOpts(
            theme='dark',
            #设置动画
            animation_opts=opts.AnimationOpts(animation_delay_update=800),#(animation_delay=1000, animation_easing="elasticOut"),
            #设置宽度、高度
            width='1500px',
            height='900px', )
        )
        .add_xaxis(list(data_bj['日期'])[0:i])
        .add_yaxis(
            series_name="",
            y_axis=list(data_bj['最高温度'])[0:i], is_smooth=True,is_symbol_show=False,
            linestyle_opts={
    
    
                   'normal': {
    
    
                       'width': 3,
                       'shadowColor': 'rgba(0, 0, 0, 0.5)',
                       'shadowBlur': 5,
                       'shadowOffsetY': 10,
                       'shadowOffsetX': 10,
                       'curve': 0.5,
                       'color': JsCode(color_js0)
                   }
               },
            itemstyle_opts={
    
    
            "normal": {
    
    
                "color": JsCode(
                    """new echarts.graphic.LinearGradient(0, 0, 0, 1, [{
                offset: 0,
                color: '#ed1941'
            }, {
                offset: 1,
                color: '#009ad6'
            }], false)"""
                ),
                "barBorderRadius": [45, 45, 45, 45],
                "shadowColor": "rgb(0, 160, 221)",
            }
        },

        )
        .add_yaxis(
            series_name="",
            y_axis=list(data_bj['最低温度'])[0:i], is_smooth=True,is_symbol_show=False,
#             linestyle_opts=opts.LineStyleOpts(color=color0[1],width=3),
            itemstyle_opts=opts.ItemStyleOpts(color=JsCode(color_js1)),
            linestyle_opts={
    
    
                   'normal': {
    
    
                       'width': 3,
                       'shadowColor': 'rgba(0, 0, 0, 0.5)',
                       'shadowBlur': 5,
                       'shadowOffsetY': 10,
                       'shadowOffsetX': 10,
                       'curve': 0.5,
                       'color': JsCode(color_js1)
                   }
               },
        )
        .set_global_opts(
            title_opts=opts.TitleOpts("北京2021年每日最高最低温度变化\n\n{}".format(title_date),pos_left=330,padding=[30,20]),
            xaxis_opts=opts.AxisOpts(type_="time",max_=x_max),#, interval=10,min_=i-5,split_number=20,axistick_opts=opts.AxisTickOpts(length=2500),axisline_opts=opts.AxisLineOpts(linestyle_opts=opts.LineStyleOpts(color="grey"))
            yaxis_opts=opts.AxisOpts(min_=y_min,max_=y_max),#坐标轴颜色,axisline_opts=opts.AxisLineOpts(linestyle_opts=opts.LineStyleOpts(color="grey"))
        )
    )
    tl.add(c, "{}".format(list(data_bj['日期'])[i]))
    tl.add_schema(
        axis_type='time',
        play_interval=100,  # 表示播放的速度
        pos_bottom="-29px",
        is_loop_play=False, # 是否循环播放
        width="780px",
        pos_left='30px',
        is_auto_play=True,  # 是否自动播放。
        is_timeline_show=False)
tl.render('1.html')

data_10 = data[(data['年份'] == 2022) & ( data['月份'] == 10)]
data_10.head()

Daily maximum temperature changes in Beijing, Shanghai, Guangzhou and Shenzhen in October

# 背景色
background_color_js = (
    "new echarts.graphic.LinearGradient(0, 0, 0, 1, "
    "[{offset: 0, color: '#c86589'}, {offset: 1, color: '#06a7ff'}], false)"
)

# 线条样式
linestyle_dic = {
    
     'normal': {
    
    
                    'width': 4,  
                    'shadowColor': '#696969', 
                    'shadowBlur': 10,  
                    'shadowOffsetY': 10,  
                    'shadowOffsetX': 10,  
                    }
                }
timeline = Timeline(init_opts=opts.InitOpts(bg_color=JsCode(background_color_js),
                                            width='980px',height='600px'))


bj, gz, sh, sz= [], [], [], []
all_max = []
x_data = data_10[data_10['城市'] == '北京']['日'].tolist()
for d_time in range(len(x_data)):
    bj.append(data_10[(data_10['日'] == x_data[d_time]) & (data_10['城市']=='北京')]["最高温度"].values.tolist()[0])
    gz.append(data_10[(data_10['日'] == x_data[d_time]) & (data_10['城市']=='广州')]["最高温度"].values.tolist()[0])
    sh.append(data_10[(data_10['日'] == x_data[d_time]) & (data_10['城市']=='上海')]["最高温度"].values.tolist()[0])
    sz.append(data_10[(data_10['日'] == x_data[d_time]) & (data_10['城市']=='深圳')]["最高温度"].values.tolist()[0])
    
    line = (
        Line(init_opts=opts.InitOpts(bg_color=JsCode(background_color_js),
                                     width='980px',height='600px'))
        .add_xaxis(
            x_data,
                  )
        
        .add_yaxis(
            '北京',
            bj,
            symbol_size=5,
            is_smooth=True,
            is_hover_animation=True,
            label_opts=opts.LabelOpts(is_show=False),
        )
  
        .add_yaxis(
            '广州',
            gz,
            symbol_size=5,
            is_smooth=True,
            is_hover_animation=True,
            label_opts=opts.LabelOpts(is_show=False),
        )
 
        .add_yaxis(
            '上海',
            sh,
            symbol_size=5,
            is_smooth=True,
            is_hover_animation=True,
            label_opts=opts.LabelOpts(is_show=False),
            
        )
 
        .add_yaxis(
            '深圳',
            sz,
            symbol_size=5,
            is_smooth=True,
            is_hover_animation=True,
            label_opts=opts.LabelOpts(is_show=False),
            
        )
        
        .set_series_opts(linestyle_opts=linestyle_dic)
        .set_global_opts(
            title_opts=opts.TitleOpts(
                title='北上广深10月份最高气温变化趋势',
                pos_left='center',
                pos_top='2%',
                title_textstyle_opts=opts.TextStyleOpts(color='#DC143C', font_size=20)),
            
            tooltip_opts=opts.TooltipOpts(
                trigger="axis",
                axis_pointer_type="cross",
                background_color="rgba(245, 245, 245, 0.8)",
                border_width=1,
                border_color="#ccc",
                textstyle_opts=opts.TextStyleOpts(color="#000"),
        ),
            xaxis_opts=opts.AxisOpts(
#                 axislabel_opts=opts.LabelOpts(font_size=14, color='red'),
#                 axisline_opts=opts.AxisLineOpts(is_show=True,
#                 linestyle_opts=opts.LineStyleOpts(width=2, color='#DB7093'))
                is_show = False
            ),
                
            
            yaxis_opts=opts.AxisOpts(
                name='最高气温',            
                is_scale=True,
#                 min_= int(min([gz[d_time],sh[d_time],sz[d_time],bj[d_time]])) - 10,
                max_= int(max([gz[d_time],sh[d_time],sz[d_time],bj[d_time]])) + 10,
                name_textstyle_opts=opts.TextStyleOpts(font_size=16,font_weight='bold',color='#5470c6'),
                axislabel_opts=opts.LabelOpts(font_size=13,color='#5470c6'),
                splitline_opts=opts.SplitLineOpts(is_show=True, 
                                                  linestyle_opts=opts.LineStyleOpts(type_='dashed')),
                axisline_opts=opts.AxisLineOpts(is_show=True,
                                        linestyle_opts=opts.LineStyleOpts(width=2, color='#5470c6'))
            ),
            legend_opts=opts.LegendOpts(is_show=True, pos_right='1%', pos_top='2%',
                                        legend_icon='roundRect',orient = 'vertical'),
        ))
    
    timeline.add(line, '{}'.format(x_data[d_time]))

timeline.add_schema(
    play_interval=1000,          # 轮播速度
    is_timeline_show=True,      # 是否显示 timeline 组件
    is_auto_play=True,          # 是否自动播放
    pos_left="0",
    pos_right="0"
)
timeline.render('2.html')

Finally, after thinking about it, it’s none of my business when I’m young. Some islands will sink if they sink. It’s better to analyze the weather in Beijing, Shanghai and Guangzhou.

Complete source code + video explanation. You can pick up the business card at the end of the article.

That’s it for today’s sharing, see you next time!