Rafael Gonzalez Gouveia gonzalezgouveia

Enseño programación para Data Science en www.gonzalezgouveia.com

gonzalezgouveia / wine_cleaning.R

Created February 5, 2019 19:47

cleaning data with dplyr

	clean_data <- winedata %>%
	select(country, points, price) %>%
	drop_na() %>% # quitando los nulos
	group_by(country) %>%
	filter(n()>2000) %>% # filtrando
	ungroup() %>%
	mutate(log_price = log(price)) # log price

gonzalezgouveia / wine_clean_plot.R

Last active February 5, 2019 19:53

cleaned plot

	set.seed(12321)
	ggplot(sample_n(clean_data, size=1000),
	aes(x=jitter(log_price, factor = 3),
	y=jitter(points, factor = 3),
	color=country)) +
	geom_point(size=2) +
	xlab('log(Precio)') +
	ylab('Puntuación') +
	ggtitle('Revisiones de vinos por país')
	ggsave('./path/to/data/clean_wine_country.png')

gonzalezgouveia / wine_all_models.R

Created February 5, 2019 20:05

modeling

	all_models <- clean_data %>%
	group_by(country) %>%
	summarise(n_obs = n(),
	b = lm(points ~ log(price))$coefficients[1],
	m = lm(points ~ log(price))$coefficients[2])

gonzalezgouveia / wine_trelliscope.R

Last active February 5, 2019 20:16

trelliscope dashboard

	path_trelliscope <- "./path/to/folder/trelliscope/"
	ggplot(clean_data,
	aes(x=log(price), y=points)) +
	geom_point() +
	geom_smooth(method=lm, se = FALSE) +
	facet_trelliscope(~ country, nrow = 1, ncol = 3,
	path = path_trelliscope,
	self_contained=TRUE)

gonzalezgouveia / mediumStoriesStats.js

Created February 14, 2019 19:17 — forked from epintos/mediumStoriesStats.js

Export Medium Stats Stores to CSV

	// Run Inspector Console in chrome and copy and paste the following code in the /stats/stories view

	function download(filename, text) {
	var pom = document.createElement('a');
	pom.setAttribute('href', 'data:text/plain;charset=utf-8,' + encodeURIComponent(text));
	pom.setAttribute('download', filename);

	if (document.createEvent) {
	var event = document.createEvent('MouseEvents');
	event.initEvent('click', true, true);

gonzalezgouveia / gapminder_hispam_read_data.R

Created February 28, 2019 06:20

	library(tidyverse)

	gdpPercap <- read_csv('/path/to/data/income_..._adjusted.csv'))
	lifeExp <- read_csv('/path/to/data/life_expectancy_years.csv'))

gonzalezgouveia / gapminder_hispam_data_prep.R

Created February 28, 2019 06:24

	# Filter for HispAm countries
	hispam_vec <- c(
	'Argentina', 'Brazil', 'Bolivia', 'Chile', 'Colombia',
	'Costa Rica', 'Cuba', 'Dominican Republic', 'Ecuador',
	'El Salvador', 'Guatemala', 'Honduras', 'Mexico',
	'Nicaragua', 'Panama', 'Paraguay', 'Peru', 'Uruguay',
	'Spain', 'Puerto Rico', 'Venezuela' )

	# dplyr::filter countries in hispam
	######

gonzalezgouveia / gapminder_hispam_result_gif.R

Last active February 28, 2019 19:16

	# install.packages(gganimate)
	library(gganimate)
	# devtools::install_github('rensa/ggflags')
	library(ggflags)
	# install.packages("gifski")
	library(gifski)

	# define animation object
	anim <- gapminder_hispam %>%
	filter(year >= 1900) %>%

gonzalezgouveia / random_read_data.py

Created March 18, 2019 20:23

	import pandas as pd
	import numpy as np

	# data from Repo https://github.com/marcboquet/spanish-names
	name_men = pd.read_csv('./hombres.csv')
	name_men_list = name_men['nombre'].tolist()
	name_women = pd.read_csv('./mujeres.csv')
	name_women_list = name_women['nombre'].tolist()

gonzalezgouveia / random_parameters.py

Created March 18, 2019 20:26