美丽的汤嵌套循环垂直输出到数据帧

Question

如何将此网络抓取（数据）的输出放入 pandas DataFrame？下面的代码只展示了一条'row'条数据，而且是竖排的。如何使 DataFrame 看起来完全像 print(data) - 即。水平的？另外，我是个新手 - 这就是 'not pythonic'。请建议如何将其变成函数（？）？我尝试了各种 append 和 concat 函数，但嵌套循环使一切都陷入错误。请帮忙。
先感谢您。

from bs4 import BeautifulSoup as bs
import urllib.request
import re
import pandas as pd

Url1 = 'http://www5.statcan.gc.ca/cimt-cicm/topNCountryCommodities-marchandises?lang=eng&chapterId=27&sectionId=0&refMonth='
Url2 = '&refYr='
Url3 = '&freq=6&countryId=999&usaState=0&provId=1&arrayId=9900000&commodityId=271111&commodityName=Natural+gas%2C+liquefied&topNDefault=10&tradeType=3'

regexQ = re.compile('Date1 Qty')
regexC = re.compile('Footnote')
regexV = re.compile('Date1 Val')
import csv
import collections

for year in range(2016, 2018):
    for month in range(0, 13):
        sauce = urllib.request.urlopen(Url1+str(month)+Url2+str(year)+Url3).read().decode('utf-8')
        soup = bs(sauce, 'lxml')
        select = soup.find_all('option', selected=True)
        for i in select:
            Month=select[4]
            Year=select[3]
            month=Month.text
            year=Year.text
            #print(year)
            #print(month, year)
        value = [x.text.strip() for x in soup.find_all("td", {"headers": regexV})]
        country = [x.text.strip() for x in soup.find_all("a", {"href": regexC})]
        quantity = [x.text.strip() for x in soup.find_all("td", {"headers": regexQ})]
        total_list = [list(x) for x in zip(country,quantity,value)]
        for item in total_list:
            h=[year , month]
            data = h + item
            print(data)
            df = pd.DataFrame(data=data)
            df.head

Answer 1

您可以使用：

from bs4 import BeautifulSoup as bs
import urllib.request
import re
import pandas as pd

Url1 = 'http://www5.statcan.gc.ca/cimt-cicm/topNCountryCommodities-marchandises?lang=eng&chapterId=27&sectionId=0&refMonth='
Url2 = '&refYr='
Url3 = '&freq=6&countryId=999&usaState=0&provId=1&arrayId=9900000&commodityId=271111&commodityName=Natural+gas%2C+liquefied&topNDefault=10&tradeType=3'

regexQ = re.compile('Date1 Qty')
regexC = re.compile('Footnote')
regexV = re.compile('Date1 Val')
import csv
import collections

#create list for all rows
data = []
for year in range(2016, 2018):
    for month in range(0, 13):
        sauce = urllib.request.urlopen(Url1+str(month)+Url2+str(year)+Url3).read().decode('utf-8')
        soup = bs(sauce, 'lxml')
        select = soup.find_all('option', selected=True)
        print (select)
        for i in select:
            month=select[4].text
            #convert year to integer 
            year=int(select[3].text)
            #print(year)
            #print(month, year)

        #replace , from numeric values, convert to integer
        value = [int(x.text.strip().replace(',','')) for x in soup.find_all("td", {"headers": regexV})]
        country = [x.text.strip() for x in soup.find_all("a", {"href": regexC})]
        quantity = [int(x.text.strip().replace(',',''))  for x in soup.find_all("td", {"headers": regexQ})]
        for item in zip(country,quantity,value):
            #append each row to list data 
            data.append([year , month] + list(item))

#create Dataframe
cols = ['year','month','country','quantity','value']
df = pd.DataFrame(data=data, columns=cols)
print (df.head())
   year     month              country   quantity     value
0  2016   January                World  279605300  69598782
1  2016   January  Trinidad and Tobago  140185912  36503857
2  2016   January               Norway  137835921  32414865
3  2016   January        United States    1583467    680060
4  2016  February                World    1255156    532059

美丽的汤嵌套循环垂直输出到数据帧

output of beautiful soup nested loop vertically into dataframe

python

beautifulsoup

nested-loops

pandas