admin管理员组文章数量:1292748
I am a master student working on my data science and visualization project with the GBD data. I wish to graph a bar graph that shows the total attributed DALYs of three different cancers by five different SDI regions, and side-by-sided to compare males and females. I wish to create a bar graph that should look something like this (for my graph the y-axis should be the five SDI locations, while in the stacked bar graph shows the number of DALYs for each cancer):
enter image description here
Here is my code for visualizing the bar graph:
ggplot(globaldata%>%filter(measure_id==2,
location_id!=1,
sex_id!=3,
age_id==22,
rei_id==110,
metric_id==1,
year==2021), aes(x=reorder(location_name, val), y=val, fill=cause_name))+
geom_bar(stat = 'identity')+
facet_wrap(~sex_name, ncol=1)+
theme_classic()+
labs(x='Location', y='Total DALYs attributable to dietary risks', title='2021', fill='GI cancer')+
coord_flip()
Here is what it looked like: enter image description here I had difficulty visualizing the side-by-side graph based on the variable 'sex_id', as when I sued position='dodge' within the geom_bar() function, it gives the side-by-side bars based on the different cancers (variable 'cause_name'). As I used facet_wrap, it showed two separated stacked bar graphs. I wish to combine them in one, as the female and male graphs are side-by-sided based on each location.
Attached is the deput( ) data of my previous data:
structure(list(measure_id = c(2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2,
2, 2, 2, 2), location_id = c(44634, 44634, 44635, 44635, 44634,
44634, 44634, 44634, 44636, 44636, 44637, 44637, 44639, 44639,
44635), location_name = c("High-middle SDI", "High-middle SDI",
"High SDI", "High SDI", "High-middle SDI", "High-middle SDI",
"High-middle SDI", "High-middle SDI", "Low-middle SDI", "Low-middle SDI",
"Low SDI", "Low SDI", "Middle SDI", "Middle SDI", "High SDI"),
sex_id = c(1, 2, 1, 2, 1, 2, 1, 2, 1, 2, 1, 2, 1, 2, 1),
sex_name = c("Male", "Female", "Male", "Female", "Male",
"Female", "Male", "Female", "Male", "Female", "Male", "Female",
"Male", "Female", "Male"), age_id = c(22, 22, 22, 22, 22,
22, 22, 22, 22, 22, 22, 22, 22, 22, 22), cause_id = c(411,
411, 411, 411, 441, 441, 414, 414, 411, 411, 411, 411, 411,
411, 441), cause_name = c("Esophageal cancer", "Esophageal cancer",
"Esophageal cancer", "Esophageal cancer", "Colon and rectum cancer",
"Colon and rectum cancer", "Stomach cancer", "Stomach cancer",
"Esophageal cancer", "Esophageal cancer", "Esophageal cancer",
"Esophageal cancer", "Esophageal cancer", "Esophageal cancer",
"Colon and rectum cancer"), rei_id = c(110, 110, 110, 110,
110, 110, 110, 110, 110, 110, 110, 110, 110, 110, 110), rei_name = c("Dietary risks",
"Dietary risks", "Dietary risks", "Dietary risks", "Dietary risks",
"Dietary risks", "Dietary risks", "Dietary risks", "Dietary risks",
"Dietary risks", "Dietary risks", "Dietary risks", "Dietary risks",
"Dietary risks", "Dietary risks"), metric_id = c(1, 1, 1,
1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1), metric_name = c("Number",
"Number", "Number", "Number", "Number", "Number", "Number",
"Number", "Number", "Number", "Number", "Number", "Number",
"Number", "Number"), year = c(1990, 1990, 1990, 1990, 1990,
1990, 1990, 1990, 1990, 1990, 1990, 1990, 1990, 1990, 1990
), val = c(460695.042829528, 145049.177193469, 195007.739437153,
53050.8912015806, 900498.811619103, 865347.948534737, 444699.585454242,
208852.577212405, 101958.484232596, 74652.7533758179, 66135.6911372065,
51358.063337564, 627554.246311558, 249829.178558246, 1059015.25574152
)), row.names = c(NA, -15L), class = c("tbl_df", "tbl", "data.frame"
))
I am a master student working on my data science and visualization project with the GBD data. I wish to graph a bar graph that shows the total attributed DALYs of three different cancers by five different SDI regions, and side-by-sided to compare males and females. I wish to create a bar graph that should look something like this (for my graph the y-axis should be the five SDI locations, while in the stacked bar graph shows the number of DALYs for each cancer):
enter image description here
Here is my code for visualizing the bar graph:
ggplot(globaldata%>%filter(measure_id==2,
location_id!=1,
sex_id!=3,
age_id==22,
rei_id==110,
metric_id==1,
year==2021), aes(x=reorder(location_name, val), y=val, fill=cause_name))+
geom_bar(stat = 'identity')+
facet_wrap(~sex_name, ncol=1)+
theme_classic()+
labs(x='Location', y='Total DALYs attributable to dietary risks', title='2021', fill='GI cancer')+
coord_flip()
Here is what it looked like: enter image description here I had difficulty visualizing the side-by-side graph based on the variable 'sex_id', as when I sued position='dodge' within the geom_bar() function, it gives the side-by-side bars based on the different cancers (variable 'cause_name'). As I used facet_wrap, it showed two separated stacked bar graphs. I wish to combine them in one, as the female and male graphs are side-by-sided based on each location.
Attached is the deput( ) data of my previous data:
structure(list(measure_id = c(2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2,
2, 2, 2, 2), location_id = c(44634, 44634, 44635, 44635, 44634,
44634, 44634, 44634, 44636, 44636, 44637, 44637, 44639, 44639,
44635), location_name = c("High-middle SDI", "High-middle SDI",
"High SDI", "High SDI", "High-middle SDI", "High-middle SDI",
"High-middle SDI", "High-middle SDI", "Low-middle SDI", "Low-middle SDI",
"Low SDI", "Low SDI", "Middle SDI", "Middle SDI", "High SDI"),
sex_id = c(1, 2, 1, 2, 1, 2, 1, 2, 1, 2, 1, 2, 1, 2, 1),
sex_name = c("Male", "Female", "Male", "Female", "Male",
"Female", "Male", "Female", "Male", "Female", "Male", "Female",
"Male", "Female", "Male"), age_id = c(22, 22, 22, 22, 22,
22, 22, 22, 22, 22, 22, 22, 22, 22, 22), cause_id = c(411,
411, 411, 411, 441, 441, 414, 414, 411, 411, 411, 411, 411,
411, 441), cause_name = c("Esophageal cancer", "Esophageal cancer",
"Esophageal cancer", "Esophageal cancer", "Colon and rectum cancer",
"Colon and rectum cancer", "Stomach cancer", "Stomach cancer",
"Esophageal cancer", "Esophageal cancer", "Esophageal cancer",
"Esophageal cancer", "Esophageal cancer", "Esophageal cancer",
"Colon and rectum cancer"), rei_id = c(110, 110, 110, 110,
110, 110, 110, 110, 110, 110, 110, 110, 110, 110, 110), rei_name = c("Dietary risks",
"Dietary risks", "Dietary risks", "Dietary risks", "Dietary risks",
"Dietary risks", "Dietary risks", "Dietary risks", "Dietary risks",
"Dietary risks", "Dietary risks", "Dietary risks", "Dietary risks",
"Dietary risks", "Dietary risks"), metric_id = c(1, 1, 1,
1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1), metric_name = c("Number",
"Number", "Number", "Number", "Number", "Number", "Number",
"Number", "Number", "Number", "Number", "Number", "Number",
"Number", "Number"), year = c(1990, 1990, 1990, 1990, 1990,
1990, 1990, 1990, 1990, 1990, 1990, 1990, 1990, 1990, 1990
), val = c(460695.042829528, 145049.177193469, 195007.739437153,
53050.8912015806, 900498.811619103, 865347.948534737, 444699.585454242,
208852.577212405, 101958.484232596, 74652.7533758179, 66135.6911372065,
51358.063337564, 627554.246311558, 249829.178558246, 1059015.25574152
)), row.names = c(NA, -15L), class = c("tbl_df", "tbl", "data.frame"
))
2 Answers
Reset to default 2Following GGplot: Two stacked bar plots side by side (not facets) one possible option would be to use two separate geom_cols to draw the stacked bars for females and males, and to manually dodge or shift the bars which requires to convert location_name to a numeric plus some additional data wrangling:
library(ggplot2)
library(dplyr, warn = FALSE)
globaldata <- globaldata |>
mutate(
location_name = reorder(location_name, val),
location_name_num = as.numeric(location_name),
sex_name_num = as.numeric(factor(sex_name)),
sex_name_num = scales::rescale(sex_name_num, to = c(-1, 1))
)
width <- .9 / 2 # Bar Width
padding <- .025 # Padding between male and female bars.
ggplot(
globaldata,
aes(
x = val,
y = location_name_num + width / 2 * sex_name_num,
fill = cause_name
)
) +
geom_col(
data = ~ filter(., sex_name == "Female"),
width = width - padding, orientation = "y"
) +
geom_col(
data = ~ filter(., sex_name == "Male"),
width = width - padding, orientation = "y"
) +
stat_summary(
data = ~ filter(., sex_name == "Female"),
aes(label = "F"),
orientation = "y",
fun = sum,
geom = "label",
fill = NA,
label.size = 0,
hjust = 0
) +
stat_summary(
data = ~ filter(., sex_name == "Male"),
aes(label = "M"),
orientation = "y",
fun = sum,
geom = "label",
fill = NA,
label.size = 0,
hjust = 0
) +
scale_y_continuous(
breaks = unique(globaldata$location_name_num),
labels = unique(globaldata$location_name)
) +
facet_wrap(~year, ncol = 1) +
theme_classic() +
labs(
x = "Location", y = "Total DALYs attributable to dietary risks",
title = "2021", fill = "GI cancer"
)
You can try the following, which facets on the SDI location and positions on the sex:
sex_data <- summarise(data,
val=sum(val), .by=c(sex_name, location_name))
ggplot(data, aes(x=reorder(sex_name, val), y=val))+
geom_bar(aes(fill=cause_name), stat = 'identity',
position = "stack", width=1)+
facet_grid(location_name~., switch="y")+
geom_text(data=sex_data, aes(label = substr(sex_name,1,1)),
hjust=-1) +
scale_y_continuous(expand = expansion(mult = c(0.01, 0.08))) +
theme_classic()+
theme(axis.text.y=element_blank(),
axis.ticks.y=element_blank(),
axis.title.y = element_blank(),
strip.text.y.left = element_text(angle = 0)) +
labs(x='Location',
y='Total DALYs attributable to dietary risks',
title='2021', fill='GI cancer')+
coord_flip()
本文标签： rQuestion regarding graphing a sidebysided stacked bar graph with ggplot2 in RStudioStack Overflow
版权声明：本文标题：r - Question regarding graphing a side-by-sided stacked bar graph with ggplot2 in RStudio - Stack Overflow 内容由网友自发贡献，该文观点仅代表作者本人， 转载请联系作者并注明出处：http://www.betaflare.com/web/1741560234a2385405.html， 本站仅提供信息存储空间服务，不拥有所有权，不承担相关法律责任。如发现本站有涉嫌抄袭侵权/违法违规的内容，一经查实，本站将立刻删除。
发表评论