在一张图中绘制多个密度图

Question

我想使用ggplot2在一张图中绘制多个密度图：

我的df如下：

structure(list(item_count = c(1L, 1L, 2L, 2L, 3L, 3L, 4L, 4L, 
5L, 5L, 6L, 8L), created_at = structure(c(18319, 18321, 18319, 
18321, 18319, 18321, 18319, 18321, 18319, 18321, 18319, 18321
), class = "Date"), frequency = c(2L, 2L, 3L, 6L, 586L, 419L, 
268L, 238L, 5L, 2L, 1L, 1L)), row.names = c(NA, -12L), class = c("grouped_df", 
"tbl_df", "tbl", "data.frame"), groups = structure(list(item_count = c(1L, 
2L, 3L, 4L, 5L, 6L, 8L), .rows = list(1:2, 3:4, 5:6, 7:8, 9:10, 
    11L, 12L)), row.names = c(NA, -7L), class = c("tbl_df", "tbl", 
"data.frame"), .drop = TRUE))

aka：

# A tibble: 12 x 3
# Groups:   item_count [7]
   item_count created_at frequency
        <int> <date>         <int>
 1          1 2020-02-27         2
 2          1 2020-02-29         2
 3          2 2020-02-27         3
 4          2 2020-02-29         6
 5          3 2020-02-27       586
 6          3 2020-02-29       419
 7          4 2020-02-27       268
 8          4 2020-02-29       238
 9          5 2020-02-27         5
10          5 2020-02-29         2
11          6 2020-02-27         1
12          8 2020-02-29         1

我尝试过：

library(tidyverse)
library(ggplot2)

df %>%
  ggplot(aes(item_count, frequency), colors=created_at) + geom_density()

遇到此错误：

Error: geom_density requires the following missing aesthetics: y

我在做什么错？

Answer 1

尝试

df %>% 
  ggplot(aes(item_count, frequency, col = created_at)) + 
  geom_density(stat = "identity")

Answer 2

您已经知道商品的投放频率，因此只需排队：

ggplot(df,aes(x=item_count,y=frequency,col=factor(created_at))) + geom_line()

您可以像这样重建后一种以得到密度，但是我认为您的频率表很奇怪：

obs = data.frame(counts=rep(df$item_count,df$frequency),
created_at=rep(df$created_at,df$frequency))

ggplot(obs,aes(x=counts,col=factor(created_at))) + geom_density(bw=0.33)

可能需要找到合适的bw

observed_data