sergiolucero

import dask.dataframe as dd

token='LUCERO'

df=dd.read_csv('s3://quantcldata/servel/CLEAN/region*.csv')
tdf=df[df.nombre.str.contains(token)]

print(tdf.compute())

sergiolucero

import requests, pickle, sys
from bs4 import BeautifulSoup
from docx import Document       
from operator import methodcaller

url_bs = lambda url: BeautifulSoup(requests.get(url).text,'html5lib')  # magic!

def recopila_acordes(artista = 'Manu Chao'):       # "Manu Chao" -> manu_chao
    fartist = '_'.join(map(methodcaller("lower"),artista.split()))   
    url = f'https://acordes.lacuerda.net/{fartist}/'

sergiolucero

import altair as alt
from vega_datasets import data

source = data.seattle_weather.url
step = 20
overlap = 1

alt.Chart(source, height=step).transform_timeunit(Month='month(date)'
).transform_joinaggregate(mean_temp='mean(temp_max)', groupby=['Month']
).transform_bin(['bin_max', 'bin_min'], 'temp_max'

sergiolucero

import folium, pandas as pd
from folium.plugins import MarkerCluster

pdf = pd.read_json('https://tinyurl.com/covid19-github')
pdf = pdf[pdf.data==pdf.data.max()]
pdf = pdf[pdf.totale_casi>0]

location = pdf.describe()[['lat','long']].loc['50%'].values
fm = folium.Map(location=location, zoom_start=6, tile='stamentoner',
               width=800, height=600)

sergiolucero

from dask.distributed import Client

def pdf2csv(fn):
    doc = fitz.open(fn)

    fw = open(fn.replace('.pdf','.csv'),'w')
    writer = csv.writer(fw)
    writer.writerow(['nombre','rut','genero','direccion'])
    for ix,page in enumerate(doc):
        t = str(page.getText().encode('latin-1'))

sergiolucero

import csv, glob, fitz

def pdf2csv(fn):

    csv_fn = fn.replace('.pdf','.csv')
    region = int(fn[1:3])   # A04101.pdf -> 4

    with open(csv_fn,'w') as fw:
        writer = csv.writer(fw)
        writer.writerow(['nombre','rut','genero','direccion',

sergiolucero

import folium
from folium.plugins import MarkerCluster

mc = MarkerCluster()
fm = folium.Map(location=[40.72,-73.98],zoom_start=5)
for _,row in ndf.iterrows():
    mc.add_child(folium.Marker(location=[row['lat'],row['lon']]))
mc.add_to(fm)
fm

sergiolucero

import pickle, time
import asyncio
import concurrent.futures
import requests

url_base='https://www.contraloria.cl/opencgrapp/geocgr/api/comunas/%05d/newobras'

def get_comunas(region_id):
    print(f'START {region_id} [{time.ctime()}]')
    region_id = int(region_id)

sergiolucero

import pandas as pd
import xml.etree.ElementTree as ET
import requests, sys

def datos(id):
    url = f'https://www.senado.cl/wspublico/tramitacion.php?boletin={id}'
    r=requests.get(url)
    root = ET.fromstring(r.text)
    if len(root):
        pdatos = [root[0][0][ix].text for ix in [2,4,7]]

sergiolucero

import requests, pandas as pd
from creds import GOOGLE_KEY

GEO_BASE = 'https://maps.googleapis.com/maps/api/geocode/json?address=%s&key=%s'

def georef(address): #1600+Amphitheatre+Parkway,+Mountain+View,+CA

    url = GEO_BASE %(address, GOOGLE_KEY) 
    res = requests.get(url).json().get('results')   #returns a list
    df = pd.DataFrame(res)