I have created an R function to create stacked boxplots so I can change the variables to create different plots. However, when it comes to printing the title (either ggtitle or labs(title)), the function cannot identify the SNP variable.
Error in dots_list(..., title = title, subtitle = subtitle, caption = caption, : object 'chrD3.54068099' not found
This is the script I have used.
boxplot_function <- function(Group,SNP) {
Group %>%
group_by(`Pheno`,{{SNP}}) %>%
count({{SNP}}) %>%
group_by(`Pheno`) %>%
mutate(`Proportion` = prop.table(`n`)) %>%
ggplot(aes(x = `Pheno`, y = `Proportion`, fill = factor({{SNP}}, levels = c("HOM_ALT", "HET", "HOM_REF"))))
geom_bar(position = "stack", stat = "identity")
labs(x = "Phenotype")
labs(fill = "Genotype")
geom_text(aes(label = round(`Proportion`, 2)), position = "stack", vjust = 2)
ggtitle({{SNP}})
}
If I remove the script's ggtitle/labs(title) part, the function works perfectly.
I have tried different variations, such as using paste/paste0, both of which do not work. When using "{{SNP}}" it just pastes "{{SNP}}" as the title.
If you can help it would be very much appreciated! Thank you in advance!!
dput:
> dput <- dput(head(Superelaxed_vcf_topvariants,5))
structure(list(Breed_ID = c("DSH_2220", "DSH_3212", "DSH_3118",
"DSH_2924", "DSH_2529"), IID = c(2220, 3212, 3118, 2924, 2529
), Pheno = c("CONTROL", "CONTROL", "CONTROL", "CONTROL", "CONTROL"
), chrA1.135833172 = c("HOM_REF", "HET", "HET", "HET", "HET"),
chrUn.19473157 = c("HOM_REF", "HOM_REF", "HET", "HOM_ALT",
"HOM_REF"), chrA3.73051425 = c("HOM_REF", "HOM_REF", "HOM_REF",
"HOM_REF", "HOM_REF"), chrB2.93712886 = c("HOM_REF", "HET",
"HOM_REF", "HET", "HOM_REF"), chrUn8.11195498 = c("HOM_REF",
"HET", "HET", "HET", "HET"), chrB3.105804049 = c("HOM_REF",
"HOM_REF", "HOM_REF", "HOM_REF", "HOM_REF"), chrD3.13004058 = c("HET",
"HOM_REF", "HOM_REF", "HET", "HOM_REF"), chrD3.14773463 = c("HET",
"HOM_REF", "HOM_ALT", "HOM_ALT", "HOM_REF"), chrD3.54068099 = c("HOM_REF",
"HOM_REF", "HOM_REF", "HOM_REF", "HOM_ALT"), chrF2.7367258 = c("HET",
"HET", "HOM_REF", "HOM_REF", "HOM_REF"), chrF2.9280651 = c("HOM_REF",
"HOM_REF", "HOM_REF", "HOM_REF", "HOM_REF")), row.names = c(NA,
5L), class = "data.frame")
Example SNP: chrF2.9280651
CodePudding user response：
edit
You should remove the levels in your factor like this:
SNP <- "chrUn.19473157"
library(ggplot2)
library(dplyr)
boxplot_function <- function(Group,SNP) {
Group %>%
group_by(`Pheno`,!!SNP) %>%
count(!!SNP) %>%
group_by(`Pheno`) %>%
mutate(`Proportion` = prop.table(`n`)) %>%
ggplot(aes(x = `Pheno`, y = `Proportion`, fill = factor(!!SNP)))
geom_bar(position = "stack", stat = "identity")
labs(x = "Phenotype")
labs(fill = "Genotype")
geom_text(aes(label = round(`Proportion`, 2)), position = "stack", vjust = 2)
ggtitle(SNP)
}
boxplot_function(data, SNP)
Created on 2022-08-27 with reprex v2.0.2
First of all, I removed the fill part in aes because your data doesn't represent all levels of your factor, but you can just remove the # in the code. You could use !! instead of {{}} and remove the {{}} in ggtitle and resulting in:
data <- structure(list(Breed_ID = c("DSH_2220", "DSH_3212", "DSH_3118",
"DSH_2924", "DSH_2529"), IID = c(2220, 3212, 3118, 2924, 2529
), Pheno = c("CONTROL", "CONTROL", "CONTROL", "CONTROL", "CONTROL"
), chrA1.135833172 = c("HOM_REF", "HET", "HET", "HET", "HET"),
chrUn.19473157 = c("HOM_REF", "HOM_REF", "HET", "HOM_ALT",
"HOM_REF"), chrA3.73051425 = c("HOM_REF", "HOM_REF", "HOM_REF",
"HOM_REF", "HOM_REF"), chrB2.93712886 = c("HOM_REF", "HET",
"HOM_REF", "HET", "HOM_REF"), chrUn8.11195498 = c("HOM_REF",
"HET", "HET", "HET", "HET"), chrB3.105804049 = c("HOM_REF",
"HOM_REF", "HOM_REF", "HOM_REF", "HOM_REF"), chrD3.13004058 = c("HET",
"HOM_REF", "HOM_REF", "HET", "HOM_REF"), chrD3.14773463 = c("HET",
"HOM_REF", "HOM_ALT", "HOM_ALT", "HOM_REF"), chrD3.54068099 = c("HOM_REF",
"HOM_REF", "HOM_REF", "HOM_REF", "HOM_ALT"), chrF2.7367258 = c("HET",
"HET", "HOM_REF", "HOM_REF", "HOM_REF"), chrF2.9280651 = c("HOM_REF",
"HOM_REF", "HOM_REF", "HOM_REF", "HOM_REF")), row.names = c(NA,
5L), class = "data.frame")
SNP <- "chrF2.9280651"
library(ggplot2)
library(dplyr)
boxplot_function <- function(Group,SNP) {
Group %>%
group_by(`Pheno`,!!SNP) %>%
count(!!SNP) %>%
group_by(`Pheno`) %>%
mutate(`Proportion` = prop.table(`n`)) %>%
ggplot(aes(x = `Pheno`, y = `Proportion`)) #, fill = factor(!!SNP, levels = c("HOM_ALT", "HET", "HOM_REF"))))
geom_bar(position = "stack", stat = "identity")
labs(x = "Phenotype")
labs(fill = "Genotype")
geom_text(aes(label = round(`Proportion`, 2)), position = "stack", vjust = 2)
ggtitle(SNP)
}
boxplot_function(data, SNP)
Created on 2022-08-27 with reprex v2.0.2
CodePudding user response：
I figured out a way - I had to change the object into a character string within my function.
boxplot_function <- function(Group,SNP) {
SNP_name <- deparse(substitute(SNP))
Group %>%
select(`Pheno`,{{SNP}}) %>%
group_by(`Pheno`) %>%
count({{SNP}}) %>%
group_by(`Pheno`) %>%
mutate(`Proportion` = prop.table(`n`)) %>%
ggplot(aes(x = `Pheno`, y = `Proportion`, fill = factor({{SNP}}, levels = c("HOM_ALT", "HET", "HOM_REF"))))
geom_bar(position = "stack", stat = "identity")
labs(x = "Phenotype")
labs(fill = "Genotype")
geom_text(aes(label = round(`Proportion`, 2)), position = "stack", vjust = 2)
ggtitle(SNP_name)
}
boxplot_function(dput,chrUn.19473157)