使用ggplot在宽格式数据框的条形图上绘制错误条
Plotting error bar on bar chart for a data frame in wide format using ggplot
您好，我想在条形图上绘制误差线。问题是因为我的数据框是宽格式的，所以我无法绘制它。数据框如下图：
structure(list(Insured_Age_Group = c(1, 2, 3, 4, 5, 6, 7), Policy_Status = c("Issuance",
"Issuance", "Issuance", "Issuance", "Issuance", "Issuance", "Issuance"
), Deposit_mean = c(3859543.73892798, 4013324.11384503, 3970469.37408863,
4405204.3601121, 4379252.01763646, 3816234.23370925, 3342252.39385489
), Deposit_n = c(31046L, 20039L, 20399L, 48677L, 30045L, 13947L,
3157L), Deposit_sd = c(2816342.35213949, 3016203.31909278, 3292567.51598225,
4345771.64693777, 4260381.02418456, 4748349.50958046, 4033440.60986956
), se_Deposit = c(31328.4343156912, 41761.74740604, 45184.1713046368,
38606.556913894, 48174.6323355127, 78805.8303265365, 140700.113248691
), Insurance_mean = c(1962975.48419977, 2003323.06714903, 2665058.97077804,
3033051.58298144, 3579542.94373979, 4338039.6868955, 4806849.35326484
), Insurance_n = c(31046L, 20039L, 20399L, 48677L, 30045L, 13947L,
3157L), Insurance_sd = c(1187550.43329336, 1065410.12671512,
1840293.78284101, 2248320.36787743, 2642040.82537531, 3128969.83541335,
3030600.81901732), se_Insurance = c(13210.075727384, 14751.455352518,
25254.5009726065, 19973.4167588603, 29875.1085068105, 51929.8475078389,
105717.653906674)), row.names = c(NA, -7L), groups = structure(list(
Insured_Age_Group = c(1, 2, 3, 4, 5, 6, 7), .rows = structure(list(
1L, 2L, 3L, 4L, 5L, 6L, 7L), ptype = integer(0), class = c("vctrs_list_of",
"vctrs_vctr", "list"))), row.names = c(NA, 7L), class = c("tbl_df",
"tbl", "data.frame"), .drop = TRUE), class = c("grouped_df",
"tbl_df", "tbl", "data.frame"))
Insured_Age_Group Policy_Status Deposit_mean Deposit_n Deposit_sd se_Deposit Insurance_mean Insurance_n Insurance_sd se_Insurance
<dbl> <chr> <dbl> <int> <dbl> <dbl> <dbl> <int> <dbl> <dbl>
1 1 Issuance 3859544. 31046 2816342. 31328. 1962975. 31046 1187550. 13210.
2 2 Issuance 4013324. 20039 3016203. 41762. 2003323. 20039 1065410. 14751.
3 3 Issuance 3970469. 20399 3292568. 45184. 2665059. 20399 1840294. 25255.
4 4 Issuance 4405204. 48677 4345772. 38607. 3033052. 48677 2248320. 19973.
5 5 Issuance 4379252. 30045 4260381. 48175. 3579543. 30045 2642041. 29875.
6 6 Issuance 3816234. 13947 4748350. 78806. 4338040. 13947 3128970. 51930.
7 7 Issuance 3342252. 3157 4033441. 140700. 4806849. 3157 3030601. 105718.
可以看出，对于Deposit_mean和Insurance_mean的每个值，我计算了se_Deposit 和 se_Insurance（标准误差）。我绘制了如下所示的平均值图：
我知道如何使用 geom_errorbar 添加错误栏，但是，我不确定如何为每个这些条形图添加相应的 se 值，因为它们在宽格式。所以基本上，我必须以某种方式将宽格式更改为长格式，以便在每个计算的 deposit_mean 和 insurance_mean 前面我有对应的标准错误
有什么帮助或建议吗？
我想我会通过转向长格式，然后转向回不同的宽格式来重塑数据：
library(dplyr)
library(tidyr)
library(ggplot2)
df2 <- df %>%
rename(Insurance_se = se_Insurance, Deposit_se = se_Deposit) %>%
pivot_longer(-c(1:2), names_sep = "_", names_to = c("type", "metric")) %>%
pivot_wider(names_from = metric, values_from = value)
这会为您提供以下格式的数据：
df2
#> # A tibble: 14 x 7
#> # Groups: Insured_Age_Group [7]
#> Insured_Age_Group Policy_Status type mean n sd se
#> <dbl> <chr> <chr> <dbl> <dbl> <dbl> <dbl>
#> 1 1 Issuance Deposit 3859544. 31046 2816342. 31328.
#> 2 1 Issuance Insurance 1962975. 31046 1187550. 13210.
#> 3 2 Issuance Deposit 4013324. 20039 3016203. 41762.
#> 4 2 Issuance Insurance 2003323. 20039 1065410. 14751.
#> 5 3 Issuance Deposit 3970469. 20399 3292568. 45184.
#> 6 3 Issuance Insurance 2665059. 20399 1840294. 25255.
#> 7 4 Issuance Deposit 4405204. 48677 4345772. 38607.
#> 8 4 Issuance Insurance 3033052. 48677 2248320. 19973.
#> 9 5 Issuance Deposit 4379252. 30045 4260381. 48175.
#> 10 5 Issuance Insurance 3579543. 30045 2642041. 29875.
#> 11 6 Issuance Deposit 3816234. 13947 4748350. 78806.
#> 12 6 Issuance Insurance 4338040. 13947 3128970. 51930.
#> 13 7 Issuance Deposit 3342252. 3157 4033441. 140700.
#> 14 7 Issuance Insurance 4806849. 3157 3030601. 105718.
然后您可以添加错误栏和您想要的任何风格调整：
ggplot(df2, aes(factor(Insured_Age_Group), mean, fill = type)) +
geom_col(position = position_dodge(width = 0.8), width = 0.6) +
geom_errorbar(aes(ymin = mean - 1.96*se, ymax = mean + 1.96*se),
width = 0.4,
position = position_dodge(width = 0.8), size = 1) +
geom_text(aes(label = scales::dollar(mean), y = mean/2), hjust = 0.5,
angle = 90, position = position_dodge(width = 0.8)) +
scale_y_continuous(labels = scales::dollar) +
scale_fill_brewer(palette = "Greens") +
labs(x = "Insured Age Group",
y = "Premium value",
title = paste("Mean value for Deposit and Insurance Annual Premium",
"for Issuance Group", sep = "\n")) +
theme_bw() +
theme(panel.border = element_blank(),
axis.line = element_line(),
plot.title = element_text(size = 18, hjust = 0.5))
您好，我想在条形图上绘制误差线。问题是因为我的数据框是宽格式的，所以我无法绘制它。数据框如下图：
structure(list(Insured_Age_Group = c(1, 2, 3, 4, 5, 6, 7), Policy_Status = c("Issuance",
"Issuance", "Issuance", "Issuance", "Issuance", "Issuance", "Issuance"
), Deposit_mean = c(3859543.73892798, 4013324.11384503, 3970469.37408863,
4405204.3601121, 4379252.01763646, 3816234.23370925, 3342252.39385489
), Deposit_n = c(31046L, 20039L, 20399L, 48677L, 30045L, 13947L,
3157L), Deposit_sd = c(2816342.35213949, 3016203.31909278, 3292567.51598225,
4345771.64693777, 4260381.02418456, 4748349.50958046, 4033440.60986956
), se_Deposit = c(31328.4343156912, 41761.74740604, 45184.1713046368,
38606.556913894, 48174.6323355127, 78805.8303265365, 140700.113248691
), Insurance_mean = c(1962975.48419977, 2003323.06714903, 2665058.97077804,
3033051.58298144, 3579542.94373979, 4338039.6868955, 4806849.35326484
), Insurance_n = c(31046L, 20039L, 20399L, 48677L, 30045L, 13947L,
3157L), Insurance_sd = c(1187550.43329336, 1065410.12671512,
1840293.78284101, 2248320.36787743, 2642040.82537531, 3128969.83541335,
3030600.81901732), se_Insurance = c(13210.075727384, 14751.455352518,
25254.5009726065, 19973.4167588603, 29875.1085068105, 51929.8475078389,
105717.653906674)), row.names = c(NA, -7L), groups = structure(list(
Insured_Age_Group = c(1, 2, 3, 4, 5, 6, 7), .rows = structure(list(
1L, 2L, 3L, 4L, 5L, 6L, 7L), ptype = integer(0), class = c("vctrs_list_of",
"vctrs_vctr", "list"))), row.names = c(NA, 7L), class = c("tbl_df",
"tbl", "data.frame"), .drop = TRUE), class = c("grouped_df",
"tbl_df", "tbl", "data.frame"))
Insured_Age_Group Policy_Status Deposit_mean Deposit_n Deposit_sd se_Deposit Insurance_mean Insurance_n Insurance_sd se_Insurance
<dbl> <chr> <dbl> <int> <dbl> <dbl> <dbl> <int> <dbl> <dbl>
1 1 Issuance 3859544. 31046 2816342. 31328. 1962975. 31046 1187550. 13210.
2 2 Issuance 4013324. 20039 3016203. 41762. 2003323. 20039 1065410. 14751.
3 3 Issuance 3970469. 20399 3292568. 45184. 2665059. 20399 1840294. 25255.
4 4 Issuance 4405204. 48677 4345772. 38607. 3033052. 48677 2248320. 19973.
5 5 Issuance 4379252. 30045 4260381. 48175. 3579543. 30045 2642041. 29875.
6 6 Issuance 3816234. 13947 4748350. 78806. 4338040. 13947 3128970. 51930.
7 7 Issuance 3342252. 3157 4033441. 140700. 4806849. 3157 3030601. 105718.
可以看出，对于Deposit_mean和Insurance_mean的每个值，我计算了se_Deposit 和 se_Insurance（标准误差）。我绘制了如下所示的平均值图：
geom_errorbar 添加错误栏，但是，我不确定如何为每个这些条形图添加相应的 se 值，因为它们在宽格式。所以基本上，我必须以某种方式将宽格式更改为长格式，以便在每个计算的 deposit_mean 和 insurance_mean 前面我有对应的标准错误
有什么帮助或建议吗？
我想我会通过转向长格式，然后转向回不同的宽格式来重塑数据：
library(dplyr)
library(tidyr)
library(ggplot2)
df2 <- df %>%
rename(Insurance_se = se_Insurance, Deposit_se = se_Deposit) %>%
pivot_longer(-c(1:2), names_sep = "_", names_to = c("type", "metric")) %>%
pivot_wider(names_from = metric, values_from = value)
这会为您提供以下格式的数据：
df2
#> # A tibble: 14 x 7
#> # Groups: Insured_Age_Group [7]
#> Insured_Age_Group Policy_Status type mean n sd se
#> <dbl> <chr> <chr> <dbl> <dbl> <dbl> <dbl>
#> 1 1 Issuance Deposit 3859544. 31046 2816342. 31328.
#> 2 1 Issuance Insurance 1962975. 31046 1187550. 13210.
#> 3 2 Issuance Deposit 4013324. 20039 3016203. 41762.
#> 4 2 Issuance Insurance 2003323. 20039 1065410. 14751.
#> 5 3 Issuance Deposit 3970469. 20399 3292568. 45184.
#> 6 3 Issuance Insurance 2665059. 20399 1840294. 25255.
#> 7 4 Issuance Deposit 4405204. 48677 4345772. 38607.
#> 8 4 Issuance Insurance 3033052. 48677 2248320. 19973.
#> 9 5 Issuance Deposit 4379252. 30045 4260381. 48175.
#> 10 5 Issuance Insurance 3579543. 30045 2642041. 29875.
#> 11 6 Issuance Deposit 3816234. 13947 4748350. 78806.
#> 12 6 Issuance Insurance 4338040. 13947 3128970. 51930.
#> 13 7 Issuance Deposit 3342252. 3157 4033441. 140700.
#> 14 7 Issuance Insurance 4806849. 3157 3030601. 105718.
然后您可以添加错误栏和您想要的任何风格调整：
ggplot(df2, aes(factor(Insured_Age_Group), mean, fill = type)) +
geom_col(position = position_dodge(width = 0.8), width = 0.6) +
geom_errorbar(aes(ymin = mean - 1.96*se, ymax = mean + 1.96*se),
width = 0.4,
position = position_dodge(width = 0.8), size = 1) +
geom_text(aes(label = scales::dollar(mean), y = mean/2), hjust = 0.5,
angle = 90, position = position_dodge(width = 0.8)) +
scale_y_continuous(labels = scales::dollar) +
scale_fill_brewer(palette = "Greens") +
labs(x = "Insured Age Group",
y = "Premium value",
title = paste("Mean value for Deposit and Insurance Annual Premium",
"for Issuance Group", sep = "\n")) +
theme_bw() +
theme(panel.border = element_blank(),
axis.line = element_line(),
plot.title = element_text(size = 18, hjust = 0.5))