相同的数据不同的展示方式，本节我们继续使用TidyTuesday 2021 第十周的项目数据集来进行数据分析；喜欢的小伙伴欢迎关注我的公众号R语言数据分析指南, 持续分享更多优质资源。后台回复关键词2021-10-2获取全文代码
本节采用流图的方式来展示数据
安装必需R包
library(tidyverse)
library(tidytuesdayR)
library(ggstream)
# remotes::install_github("davidsjoberg/ggstream")
加载数据
tues_data <- tidytuesdayR::tt_load(2021, week = 10)
youtube <- tues_data$youtube
数据可视化
通过case_when函数更改变量的名称，并将变量转换为因子
youtube %>%
select(year, brand, funny:use_sex, like_count, view_count) %>%
pivot_longer(cols = funny:use_sex, names_to = "type") %>%
filter(value == T) %>%
mutate(type = factor(case_when(type == "animals" ~ "Contains animals",
type == "celebrity" ~ "Contains celebrity",
type == "danger" ~ "Contains danger",
type == "funny" ~ "Contains humor",
type == "patriotic" ~ "Patriotic",
type == "show_product_quickly" ~ "Shows product quickly",
type == "use_sex" ~ "Uses sexuality")),
brand = ifelse(brand == "Hynudai", "Hyundai", brand),
brand = factor(brand, levels = c("Coca-Cola", "Pepsi", "Bud Light", "Budweiser", "Doritos", "NFL", "E-Trade", "Toyota", "Hyundai", "Kia"), ordered = T)) %>%
group_by(year, brand, type) %>%
summarise(rel_likes = sum(like_count, na.rm = T)/sum(view_count, na.rm = T)) %>%
ggplot(aes(year, rel_likes, fill = brand)) +
geom_stream(extra_span = 0.1,
true_range = "both") +
coord_flip() +
labs(title = "Most liked Superbowl commercial types,
by brand, from 2000-2020",
fill = "")+
scale_x_continuous(limits = c(2000, 2020),
expand = c(0, 0), breaks = seq(2000, 2020, 1)) +
scale_fill_manual(values = c("#9d0208", "#dc2f02", "#f48c06", "#ffba08", "#d9ed92", "#f1faee",
"#2a9d8f", "#0077b6", "#023e8a", "#03045e")) +
facet_wrap(vars(type), nrow = 1) +
theme(panel.grid.minor = element_blank(),
panel.grid.major.x = element_blank(),
panel.background = element_rect(fill = NA),
axis.title = element_blank(),
axis.line.x = element_blank(),
axis.text.x = element_blank(),
axis.ticks.x = element_blank(),
strip.background = element_rect(fill = NA),
text = element_text(size = 15),
legend.spacing.x = unit(0.25, "cm"),
plot.title = element_text(face = "bold", size = 20),
plot.margin = unit(c(1,1,1,1),"cm"))
