ggplot why does adding a new element overwrite colour inside loop but not outside

Question

I am writing a function to plot n series of data. My idea was to loop through each series and adding a new gg_smooth for each loop. It works when I "loop" by hand but inserting it into an actual loop overwrites the colour aestetic.

Data I am working with, the idea is to be able to have n number of columns:

data

Using the following lines i get the desired result:

gene_list <- c("tetA", "tet.W.")
gg <- ggplot()

gg <- gg + geom_smooth(data=df_analysis_summed, 
              aes(x=as.Date(dato), y=!!sym(gene_list[1]), linetype = oua_2, colour = gene_list[1] ),
              method="auto", se=F)

gg <- gg + geom_smooth(data=df_analysis_summed, 
              aes(x=as.Date(dato), y=!!sym(gene_list[2]), linetype = oua_2, colour = gene_list[2] ),
              method="auto", se=F)

gg + labs(colour = "gene")

Desired outcome

I then try to add the functionality to a loop:

plot_genes_scat_smooth <- function (df,gene_list) {
  plot <- ggplot()

  for (gene_index in 1:length(gene_list)) {
    print(gene_index)
    print(gene_list[gene_index])
    plot <- plot +
      geom_smooth(data=df, aes(x=as.Date(dato), y=!!sym(gene_list[gene_index]), linetype=oua_2, colour = gene_list[gene_index]), method="auto", se=F)#+
      #geom_point(data=df,aes(x=as.Date(dato), y=!!sym(gene), colour = gene, shape = oua_2))
  }
  plot
}

genes = c("tetA", "tet.W.")
plot_genes_scat_smooth(df_analysis_summed,gene_list = genes)

Using the function I get the following result:

Wrong result from loop/function

It would seem that the colour aes of the first line is overwritten by the second call when doing trying to implement it as a function. How can that be?

Hey, i've got a similar question a while before. You might check here. Maybe one of those answers are helping you out. I guess the functionallity is the same — mischva11
– mischva11, Commented Mar 4, 2020 at 13:48

Dave2e · Accepted Answer · 2020-03-05 13:43:34Z

Instead of trying to add multiple lines, one-by-one onto the plot. Ggplot2 works best with the data in a long format. In this case the data is pivoted so that there one column for gene type and one column for the corresponding value.

#create some data
set.seed(1)
dato<-seq.Date(as.Date("2018-04-25"), length.out = 14, by="1 day")
oua_2<-rep(c(0, 1), 7)
tetA<-rnorm(14, 0.04, 0.02)
tet.W<-rnorm(14, 0.2, 0.02)

df_analysis_summed <- data.frame(dato, oua_2, tetA, tet.W)

#convert the data frame to long
library(tidyr)
df_analysis_long <- df_analysis_summed %>% pivot_longer(starts_with("tet"), names_to = "genes", values_to = "value")

#function to plot
plot_genes_scat_smooth_long <- function (df) {
  plot <- ggplot()
  plot <- plot +
    geom_smooth(data=df_analysis_long, aes(x=as.Date(dato), y=value, linetype=as.factor(oua_2), colour = genes), method="auto", se=F)

  plot
}

plot_genes_scat_smooth_long(df_analysis_long)

Collectives™ on Stack Overflow

ggplot why does adding a new element overwrite colour inside loop but not outside

1 Answer 1

Comments

Your Answer

Linked

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

Comments

Your Answer

Sign up or log in

Post as a guest

Linked

Related