Bdtjtk

I'm trying to make a boxplot with X being detections of 3 types of a certain bird genus(Vermivora): two species and hybrids of the two. There were also occasions when none of the 3 were detected, yielding a 4th level of X, "none", in addition to "gwwa" "bwwa" and "hybrid".

I only want to plot "gwwa" and "bwwa", and I want "gwwa" to be yellow and "bwwa" to be lightblue.

I can't generate a boxplot without either a third box "NA", and when I try to omit NA then there is still a third box "FALSE".

The recurring error message I receive when making attempts to avoid those two outcomes is this:

Error: Aesthetics must be either length 1 or the same as the data (3): fill

Here is what I've written:

library(tidyverse)

VERM <- read.csv("C:/1Zack/GWWA Map Stuff/GWWA_bySpecies.csv")

Vermivora <- factor(VERM$Species,levels(VERM$Species)[c(2,1)])

Road_Proximity <- VERM$RoadCount_2km



ggplot(data=VERM, mapping = aes(Vermivora, y=Road_Proximity, fill=Vermivora))+

geom_boxplot()+

scale_fill_manual(values=c("yellow", "lightblue"))

Sorry, this is my first post, and I can't figure out how to include a screenshot and I can't figure out how to put each line of my code on its own line without a full empty line between them.

I am just guessing how your data looks like (since you didn't provide it) so I made up some VERM data.frame in the example below. The main idea is that you should filter for what values you need in Vermivora variable (column).

library(tidyverse)

# Some made up data

VERM <- data.frame(Vermivora = c("gwwa", "bwwa", "hybrid", NA),

                   Road_Proximity = sample(4*10))

I presume this is similar to what you got based on the description of your problem:

ggplot(VERM, aes(Vermivora, Road_Proximity, fill = Vermivora)) + 

  geom_boxplot()

Solution with dyplr

Filter for the values that you need in the variable Vermivora.

VERM %>%

  filter(Vermivora %in% c("gwwa", "bwwa")) %>% 

  ggplot(aes(x = Vermivora, y = Road_Proximity, fill = Vermivora)) + 

  geom_boxplot() +

  scale_fill_manual(values = c("gwwa" = "yellow", 

                               "bwwa" = "lightblue"))

If you need a certain order on the OX axis, then order the factor Vermivora.

VERM %>%

  filter(Vermivora %in% c("gwwa", "bwwa")) %>% 

  mutate(Vermivora = ordered(x = Vermivora, 

                             levels = c("gwwa", "bwwa"))) %>% 

  ggplot(aes(x = Vermivora, y = Road_Proximity, fill = Vermivora)) + 

  geom_boxplot() +

  scale_fill_manual(values = c("gwwa" = "yellow", 

                               "bwwa" = "lightblue"))

Solution with data.table

Using the more concise syntax of data.table:

library(data.table)

library(magrittr) # for piping with %>% (not only dyplr use it; 

# pipeline is a Unix trait not a dyplr trait)



setDT(VERM) # converts to data.table from data.frame



VERM[Vermivora %in% c("gwwa", "bwwa")] %>% 

  ggplot(aes(x = Vermivora, y = Road_Proximity, fill = Vermivora)) + 

  geom_boxplot() +

  scale_fill_manual(values = c("gwwa" = "yellow", 

                               "bwwa" = "lightblue"))



# If you need to order the factor Vermivora:

VERM[Vermivora %in% c("gwwa", "bwwa")] %>% 

  .[, Vermivora := ordered(x = Vermivora, levels = c("gwwa", "bwwa"))] %>% 

  ggplot(aes(x = Vermivora, y = Road_Proximity, fill = Vermivora)) + 

  geom_boxplot() +

  scale_fill_manual(values = c("gwwa" = "yellow", 

                               "bwwa" = "lightblue"))

Note that when addressing a question, is almost always needed to see some data related to your problem. You can try that by posting the output of one of the following lines:

dput(VERM) # the entire data

dput(head(VERM, 20)) # the first 20 rows

dput(VERM[sample(x = nrow(VERM), size = 20),]) # a sample of 20 rows

# Or make up some data like I did in the above example.