reading in CSVs according to ui input selection - r

Question

I'm having trouble reading in .csv files dynamically based on ui input.
I have observations of precipitation and temperature data for different areas (ID'd here as zone). For a subset of those same areas, I also have frequency data of elevation observations. The app is meant as a CDF plotter of these three measurements for a region with many areas.
The problem is the elevation frequency data is much higher resolution than the precipitation and temperature data (together as p_t) and my code is too inefficient for good Shiny performance whenever users select elevation data (elev).
Instead of making one globally available filterable data.frame of all the data, I'm hoping to have Shiny go and grab individual .csvs of the frequency elevation data on the fly for just the area(s) that are selected in input$zone (via SelectizeInput). Once they're read in, uncount() so they're in observation format, then bundle together into one data.frame data() with any precipitation or temperature (observation) data that's selected, for ggplot's handy stat_ecdf() function with aesthetics set to data()$param and data()$zone.
Does this sound like an OK approach? If so, could you help me? In summary the main need is, if the parameter elev is selected, have Shiny find and read in those elevation .csvs whose file names match the areas selected. Thanks in advance for any help.
library(tidyverse)
library(shiny)
library(shinydashboard)
# generate sample observation data - precipitation and temperature
# not many megabytes so will just read in all at once at startup as one .csv
zone <- c(rep("abcde", 6), rep("fghij", 6), rep("klmno", 6),
rep("pqrst", 6))
set.seed(1)
val <- rnorm(24, 12, 18)
param <- rep(c("p", "t"), 12)
p_t <- data.frame(zone, val, param, stringsAsFactors = FALSE)
#head(p_t)
# zone val param
#1 abcde 0.7238314 p
#2 abcde 15.3055798 t
#3 abcde -3.0413150 p
#4 abcde 40.7150544 t
#5 abcde 17.9311399 p
#6 abcde -2.7684309 t
# generate sample elevation frequency data with many more observations -
# need to be stored as individual .csvs, too big to read in, uncount, and
# filter by input - too slow
# just want to bind and uncount as they're selected by user
setwd(./elevdata) # separate folder from which to pick out the elev data .csv in the main wd
# and prevent reading in of other app .csv data
val <- c(503, 506, 513, 689)
count <- c(32282, 53172, 45237, 34534)
data.frame(val, count) %>% mutate(zone = "abcde", param = "elev") %>%
write_csv("abcde.csv")
val <- c(-36, -39, -51, -98)
count <- c(52220, 5175, 299237, 100034)
data.frame(val, count) %>% mutate(zone = "fghij", param = "elev") %>%
write_csv("fghij.csv")
val <- c(2, 7, 13, 110)
count <- c(99222, 883172, 114237, 8347633)
data.frame(val, count) %>% mutate(zone = "klmno", param = "elev") %>%
write_csv("klmno.csv")
#only a subset of p_t zones have elev data - variable not currently used
#zoneswithelevdata <- list.files(pattern = "*.csv$")
#zoneswithelevdata <- gsub(".csv", "", zoneswithelevdata)
#shiny app using the above sample data
shinyApp(
ui = fluidPage(
sidebarLayout(sidebarPanel(
selectizeInput(
"zone", "zone", choices = unique(p_t$zone),
selected = c("a"),
multiple = TRUE),
checkboxGroupInput("param", "parameter",
choices = c("elev", "p", "t"), selected =c("elev", "p"))
),
mainPanel(
tabsetPanel(position=c("right"),
tabPanel(strong("static cdf"),
br(),
plotOutput("reg_plot", height = "750px")) )))
),
server = function(input, output) {
# elev_csv_counts_tobind <- reactive({
# if `elev` parameter is checked:
#* read_csv() of the csv(s) with csv file name %in% input$zone **
# * and bind together
#})
data <- reactive({
p_t_e <- p_t %>%
#first subset p_t by the zone(s) and param(s) selected
filter(param %in% input$param,
zone %in% input$zone) %>%
#now attach and uncount the elevation data
bind_rows({elev_csv_counts_tobind %>%
uncount(count)})
})
output$reg_plot <- renderPlot({
ggplot(data(), aes(val, color = param, linetype = zone)) +
labs(y = "proportion of total", x = NULL) +
stat_ecdf(pad = FALSE) + coord_flip()
})
}
)