minijumboscraper

jumbito.py

import requests, pandas as pd
from bs4 import BeautifulSoup

url_base = 'https://nuevo.jumbo.cl/%s'

def parse(what):
    print(what)
    bs = BeautifulSoup(requests.get(url_base %what).text, 'lxml')
    producto = [pt.text.split(chr(10))[1] for pt in bs.find_all('div',attrs={'class':'product-item__info'})]
    marca = [pt.text.split(chr(10))[2] for pt in bs.find_all('div',attrs={'class':'product-item__info'})]
    precios = [float(p.text.replace('$','').replace('.','').replace(',','.')) 
            for p in bs.find_all('span',attrs={'class':"product-prices__value product-prices__value--best-price"})]
    df = pd.DataFrame({'producto': producto, 'marca':marca, 'precio': precios})
    print(df.head())
    
for what in ['abarrotes','coctel','conservas']:
    parse(what)

plot above was as simple as defining familia:

import matplotlib.pyplot as plt
bdf['familia'] = [p.split()[0] for p in bdf['producto']]
bdf=bdf[bdf.familia.isin(['Vino','Shampoo','Crema','Desodorante','Alimento','Queso'])]
bdf.boxplot('precio',by='familia')
plt.show()