Displaying only those x values selected by the user - r

Question

I'm new to R. I have a large dataset that I want the user to be able to select the x values plotted on a graph. To make it easier, I've done the same thing using the mpg dataset:
library(shiny)
ui <- fluidPage(
selectInput(
inputId= "manuf",
label= "Manufacturer",
choices= mpg$manufacturer,
multiple= TRUE
),
plotOutput("graph1")
)
server <- function(input, output) {
output$graph1 <- renderPlot({
ggplot() +
geom_point (
mapping = aes (
x= input$manuf,
y= ???
)
)
})
}
shinyApp(ui = ui, server = server)
I can't for the life of me figure out what the correct syntax is for the 'y' input. I have been googling my heart out and can't figure it out, and I'm sure it's relatively simple. I want it to only output the data for whatever you've selected in the drop down.
putting in y= mpg$hwy shows ALL hwy datapoints when one manufacturer is selected and throws an error ("Aesthetics must be either length 1 or the same as the data") with more. I think the errors are self-explanatory, but that doesn't help me figure out the correct code for 'y'. Ideas? Thanks in advance.

Answer 1

The aesthetic mappings for ggplot (like aes(x = ...)) should be column names, but you aren't giving the user a choice of column names, you give the user the choice of manufacturer values---which correspond to rows. If you want the user to select certain rows to plot based on the manufacturer, you should subset/filter the data that you give to ggplot, perhaps like this:
library(shiny)
library(ggplot2)
ui <- fluidPage(
selectInput(
inputId = "manuf",
label = "Manufacturer",
choices = mpg$manufacturer,
multiple = TRUE
),
plotOutput("graph1")
)
server <- function(input, output) {
output$graph1 <- renderPlot({
ggplot(data = mpg[mpg$manufacturer %in% input$manuf, ]) +
geom_point (
mapping = aes (
x = manufacturer,
y = hwy
)
)
})
}
shinyApp(ui = ui, server = server)

Answer 2

Let's forget about Shiny for a moment and focus on how you would filter a dataset for plotting with ggplot(). The tidyverse approach is to use dplyr::filter:
library(dplyr)
library(ggplot2)
mpg %>%
filter(manufacturer == "audi") %>%
ggplot(aes(manufacturer, hwy)) +
geom_point()
So your server function would look something like this (untested):
server <- function(input, output) {
output$graph1 <- renderPlot({
mpg %>%
filter(manufacturer == input$manuf) %>%
ggplot(aes(manufacturer, hwy)) +
geom_point()
)}
}