# paquetes ----------------------------------------------------------------
library(tidyverse)
library(glue)
library(ggtext)
library(showtext)
# fuente ------------------------------------------------------------------
# colores, Nord
c1 <- "#AD8CAE"
c2 <- "#EDDAEB"
c3 <- "#222B4C"
c4 <- "white"
c5 <- "black"
# texto gral
font_add_google(name = "Ubuntu", family = "ubuntu")
# números, porcentajes, cantidades
font_add_google(name = "Victor Mono", family = "victor", db_cache = FALSE)
# eje horizontal, temporadas
font_add_google(name = "Tilt Prism", family = "tilt", db_cache = FALSE)
# íconos
font_add("fa-brands", "icon/Font Awesome 6 Brands-Regular-400.otf")
font_add("fa-solids", "icon/Font Awesome 6 Free-Solid-900.otf")
showtext_auto()
showtext_opts(dpi = 300)
# caption
fuente <- glue("Datos: <span style='color:{c3};'><span style='font-family:mono;'>{{<b>tidytuesdayR</b>}}</span> semana 33. *spam* dataset en {{kernlab}}</span>")
autor <- glue("Autor: <span style='color:{c3};'>**Víctor Gauto**</span>")
icon_twitter <- glue("<span style='font-family:fa-brands;'></span>")
icon_github <- glue("<span style='font-family:fa-brands;'></span>")
usuario <- glue("<span style='color:{c3};'>**vhgauto**</span>")
sep <- glue("**|**")
mi_caption <- glue("{fuente}<br>{autor} {sep} {icon_github} {icon_twitter} {usuario}")
# datos -------------------------------------------------------------------
browseURL("https://github.com/rfordatascience/tidytuesday/blob/master/data/2023/2023-08-15/readme.md")
spam <- readr::read_csv('https://raw.githubusercontent.com/rfordatascience/tidytuesday/master/data/2023/2023-08-15/spam.csv')
# me interesan los elementos relativos al dinero: money, dollar, n000
# cómo se diferencian los e-mails que son o NO son spam
# tiene que tener los elementos: porcentajes > 0
d <- spam |>
select(dollar, money, n000, yesno) |>
pivot_longer(cols = -yesno, names_to = "param", values_to = "valor") |>
filter(valor != 0)
# figura ------------------------------------------------------------------
# líneas de referencia horizontales
lineas_tbl <- tibble(
y = c(120, 60, 80, 40, 80, 40),
param = c("dollar", "dollar", "money", "money", "n000", "n000")
)
# cantidad de datos, para agregar en las esquinas
d_cantidad <- d |>
count(yesno, param) |>
inner_join(
slice_max(
lineas_tbl,
order_by = y,
n = 3),
by = join_by(param))
# cantidad de e-mails que SÍ son spam
d_yes <- d_cantidad |>
filter(yesno == "y") |>
mutate(label = glue("n = {n}"))
# cantidad de e-mails que NO son spam
d_no <- d_cantidad |>
filter(yesno == "n") |>
mutate(label = glue("n = {n}"))
# etiquetas para los strips
etiq <- c(
dollar = "**$**",
money = "**money**",
n000 = "**000**"
)
# colores para representar SÍ/NO es spam
yesno_colores <- c(SÍ = alpha("white", .9), NO = alpha("black", .9))
# título y subtítulo
# palabra SPAM en blanco y negro
spam_blanco <- "<b style='color:white;'>SPAM</b>"
spam_negro <- "<b style='color:black;'>SPAM</b>"
mi_title <- glue("{spam_blanco}{spam_negro}{spam_blanco}{spam_negro}")
mi_subt <- glue(
"¿Qué caracteriza al **spam**? Por lo visto, hablar de plata.
Al analizar {nrow(d)} e-mails, se
contabilizó el porcentaje de términos relativos al dinero:
el signo **$**, la palabra **money** y tres ceros **000**.
Se indican las cantidades (n) de e-mails en cada panel.")
# figura
g <- d |>
mutate(yesno = if_else(yesno == "y", "SÍ", "NO")) |>
mutate(yesno = factor(yesno, levels = c("SÍ", "NO"))) |>
ggplot(aes(valor, fill = yesno, color = yesno)) +
# horizontales
geom_hline(
data = lineas_tbl, aes(yintercept = y),
linetype = 3, color = c2) +
# cantidad, eje vertical, línea horizontal
geom_text(
data = lineas_tbl, aes(x = 0, y = y, label = y), inherit.aes = FALSE,
hjust = 0, vjust = 0, nudge_y = 1, family = "victor", color = c2, size = 3) +
# línea de frecuencia
geom_freqpoly(
binwidth = .075, alpha = .9) +
# cantidades, n, YES
geom_text(
data = d_yes, aes(2, y, label = label), inherit.aes = FALSE,
hjust = 0, vjust = 0, color = yesno_colores["SÍ"], nudge_y = +1,
family = "victor", size = 4, fontface = "bold") +
# cantidades, n, NO
geom_text(
data = d_no, aes(2, y, label = label), inherit.aes = FALSE,
hjust = 0, vjust = 1, color = yesno_colores["NO"], nudge_y = -1,
family = "victor", size = 4, fontface = "bold") +
# facetas
facet_wrap(vars(param), scales = "free", labeller = as_labeller(etiq)) +
# escalas
scale_x_log10(
breaks = c(.01, .1, 1, 10),
labels = c("0,01%", "0,1%", "1%", "10%")) +
scale_y_continuous(expand = c(0, 0)) +
scale_color_manual(
values = yesno_colores,
labels = c("<b style='color:white;'>SÍ</b>", "<b style='color:black;'>NO</b>")) +
coord_cartesian(clip = "off") +
labs(
title = mi_title,
subtitle = mi_subt,
x = "Porcentaje del total de caracteres del e-mail<br><span style='font-size:10pt'>*escala logarítmica*</span>",
y = "Cantidad",
color = "¿Es spam? ",
caption = mi_caption) +
guides(
color = guide_legend(
override.aes = list(linewidth = 4))
) +
theme_minimal() +
theme(
aspect.ratio = 1,
plot.background = element_rect(
fill = c1, color = c2, linewidth = 3),
plot.margin = margin(15, 20, 10, 15),
plot.title = element_markdown(size = 76, family = "tilt"),
plot.subtitle = element_textbox_simple(
family = "ubuntu", size = 14, color = c3),
plot.caption = element_markdown(
hjust = 1, color = c2, margin = margin(10, 0, 0, 0)),
axis.text.x = element_text(color = c2, family = "victor"),
axis.text.y = element_blank(),
axis.title.x = element_markdown(
color = c3, margin = margin(10, 0, 5, 0), family = "ubuntu", size = 17),
axis.title.y = element_markdown(color = c3, family = "ubuntu", size = 17),
axis.ticks.x = element_line(color = c2),
panel.spacing.x = unit(1, "cm"),
panel.grid = element_blank(),
strip.text = element_markdown(
color = c3, family = "ubuntu", size = 15,
margin = margin(5, 0, 30, 0)),
legend.position = "top",
legend.margin = margin(10, 0, 5, 0),
legend.title = element_text(family = "ubuntu", color = c3, size = 14),
legend.text = element_markdown(family = "ubuntu", size = 12)
)
# guardo
ggsave(
plot = g,
filename = "2023/semana_33/viz.png",
width = 30,
height = 19.17,
units = "cm")
# abro
browseURL("2023/semana_33/viz.png")
