library(tidyverse)
library(readxl)
data <- read_excel("data/top6.xlsx")
library(forcats)
data %>% ggplot(aes(fct_infreq(brand), fill = brand)) +
geom_bar() +
geom_text(stat = "count", aes(label = after_stat(count)), vjust = -0.5)+
scale_y_continuous(limits = c(0,100))+
guides(x = guide_axis(angle = 45)) +
labs(x = "品牌", fill = "品牌") +
theme_bw() +
theme(text = element_text(size = 15),
legend.position = "bottom",
legend.text = element_text(size = 10))广州6大品牌奶茶店消费价格分析
第3-4章
Li Zongzhang
2024-05-16
1 数据和变量
数据来源：大众点评网，广州六大品牌奶茶店
样本容量：336
数据来源：大众点评网，广州6大奶茶品牌
零售价格: retail.price
团购价格：group.price
品牌： brand
区位：area
市区：荔湾，越秀，天河，海珠
近郊：白云，番禺，黄埔
远郊：花都，从化，南沙，增城
2 图形工具
2.1 帕累托图
2.2 分组条形图
2.3 分组直方图
2.4 分组箱线图
2.5 散点图
2.1 帕累托图
2.2 分组条形图
data %>%
group_by(area, brand) %>%
summarise(count = n(), .groups = "drop") %>%
mutate(brand_order = paste(brand, area, rank(-count), sep = "_")) %>%
ggplot(aes(reorder(brand_order, -count),
count, fill = brand)) +
geom_col()+
facet_wrap(~ area, scales = "free_x", ncol = 3) +
geom_text(aes(label = count), vjust = -0.5) +
scale_x_discrete(labels = function(x) gsub("_.+$", "", x)) + # 移除排序编号
scale_y_continuous(limits = c(0, 42)) +
labs(x = "品牌", y = "数量") +
theme(axis.text.x = element_text(angle = 42, hjust = 1),
legend.position = "none")
2.3 分组直方图
#按单个定性变量分组
data %>%
ggplot(aes(retail.price, fill = brand))+
geom_histogram(breaks = seq(6, 38, 2))+
facet_wrap(~brand, ncol = 2) +
scale_y_continuous(limits = c(0,60))+
scale_fill_brewer(palette = "Set1") +
labs(title = "零售价直方图",
x = "零售价",
y = "频数") +
scale_x_continuous(breaks = seq(6, 36, 4),
labels = seq(6, 36, 4)) +
theme_bw() +
theme(text = element_text(size = 15),
legend.position = "bottom",
legend.text = element_text(size = 10))
零售价格：喜茶>茶百道>TANNING>CoCo都可>书亦烧仙草>益禾堂
2.4 分组箱线图
零售价格：喜茶>茶百道>TANNING>CoCo都可>书亦烧仙草>益禾堂
2.5 散点图
无论在市区、近郊、远郊，同一品牌的团购价格相同。
3 报告描述性统计量
全样本
分品牌零售价格比较
分区域零售价格比价
3.1 全样本
library(psych)
data %>%
select(retail.price, groupbuy.price, comment.num) %>%
describe() %>%
select(n, mean, sd, median, min, max, skew, kurtosis) n mean sd median min max skew kurtosis
retail.price 336 15.60 5.26 15.0 7 35 1.53 1.72
groupbuy.price 301 10.85 4.00 10.5 6 18 0.22 -1.15
comment.num 336 564.28 1027.57 281.0 13 10598 5.24 36.47表格中添加分布图像
{gtExtras}
表格中添加分布图像
{gtExtras}
retail_group <- data %>%
select(brand, retail.price) %>%
group_by(brand) %>%
summarize(mean = mean(retail.price) %>% round(3),
median = median(retail.price)%>% round(3),
sd = sd(retail.price)%>% round(3),
retail_t = list(retail.price))
retail_group %>%
arrange(desc(mean)) %>%
gt() %>%
gt_plt_dist(retail_t,
type = "histogram",
line_color = "purple",
fill_color = "green",
bw = 1)| brand | mean | median | sd | retail_t |
|---|---|---|---|---|
| 喜茶 | 27.930 | 28 | 1.818 | |
| 茶百道 | 15.603 | 16 | 1.154 | |
| TANING手挞柠檬茶 | 15.023 | 15 | 1.067 | |
| CoCo都可 | 14.391 | 14 | 2.013 | |
| 书亦烧仙草 | 13.415 | 13 | 1.770 | |
| 益禾堂 | 10.067 | 10 | 1.286 |
3.2 分品牌零售价格比较
data %>%
group_by(brand) %>%
summarize(n(),
min = min(retail.price),
q1 = quantile(retail.price, 0.25),
median = median(retail.price),
mean = mean(retail.price),
q3 = quantile(retail.price, 0.75),
max = max(retail.price),
sd = sd(retail.price)) %>%
arrange(desc(median))# A tibble: 6 × 9
brand `n()` min q1 median mean q3 max sd
<chr> <int> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl>
1 喜茶 43 24 27 28 27.9 29 35 1.82
2 茶百道 58 13 15 16 15.6 16 19 1.15
3 TANING手挞柠檬茶 44 13 14.8 15 15.0 16 18 1.07
4 CoCo都可 64 11 13 14 14.4 15 23 2.01
5 书亦烧仙草 82 10 12 13 13.4 14.8 21 1.77
6 益禾堂 45 7 9 10 10.1 11 13 1.29零售价格：喜茶>茶百道>TANNING>CoCo都可>书亦烧仙草>益禾堂
3.3 分区域零售价格比较
library(psych)
data %>%
group_by(area) %>%
summarize(n(),
min = min(retail.price),
q1 = quantile(retail.price, 0.25),
median = median(retail.price),
mean = mean(retail.price),
q3 = quantile(retail.price, 0.75),
max = max(retail.price),
sd = sd(retail.price))%>%
arrange(desc(median))# A tibble: 3 × 9
area `n()` min q1 median mean q3 max sd
<chr> <int> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl>
1 市区 168 8 13 15 16.5 16.2 35 5.90
2 近郊 138 7 13 14 14.7 16 30 4.20
3 远郊 30 8 11 14 14.5 16 28 5.06零售价格：市区>近郊>远郊
4 研究结论
门店数量：书亦烧仙草>CoCo都可>茶百道>益禾堂>TANNING>喜茶
零售价格：喜茶>茶百道>TANNING>CoCo都可>书亦烧仙草>益禾堂
团购价格：喜茶>TANNING>茶百道>CoCo都可>益禾堂>书亦烧仙草
无论在市区、近郊、远郊，同一个品牌的团购价格相同。
零售价格：市区>近郊>远郊
5 经验分享
数据来源
清洗数据
R代码
{gtExtras}
提炼结论
Thanks for your attention！
广州6大品牌奶茶店消费价格分析 第3-4章 Li Zongzhang 2024-05-16