This one has me really going around in circles.
I am working on an R script that loads a dataframe and uses fields from the dataframe to populate a hierarchical set of selectizeInput. E.g. each of the inputs represent a subset of what is in the previous. Each SubRegion contains multiple LCC’s, Each LCC contains multiple ENB’s, and so on.
When the user select a value in any of the inputs, that value will used to filter the dataframe and all of the other selectizeInputs need to be updated from the filtered data.
It seems to work fine for the first input (SubRegionInput) but every time I try to get it to respond to and/or filter by any of the others (e.g. add input$LCCInput to the observe block) they get populated for a few seconds and then go blank.
I suspect the answer is quite simple and/or I am doing something really dumb, but I am a total hack with no formal R training so am probably missing something quite basic (if so sorry).
Below is a partial chunk of code (sorry I can’t include it all but this is for work and I can’t share the details of what I am doing).
NOTES
The current outputs are just so I can see what is going on while I develop this portion of the code.
I know right now it is only set up to filter on the one value…everything I have tried to do it on more has failed so I included the most functional code I have so far.
ui <- fluidPage(
# Application title
titlePanel("KPI DrillDown"),
# Sidebar with a slider input for number of bins
fluidRow(
selectizeInput("SubRegionInput", "SubRegion", SubRegionList ,selected = NULL, multiple = TRUE),
selectizeInput("LCCInput", "LCC", LCCList,selected = NULL, multiple = TRUE),
selectizeInput("ENBIDInput", "ENBID", ENBIDList,selected = NULL, multiple = TRUE),
selectizeInput("SiteNumInput", "SiteNumber", SiteNumberList,selected = NULL, multiple = TRUE),
selectizeInput("SiteNameInput", "SiteName", SiteNameList,selected = NULL, multiple = TRUE),
selectizeInput("LNCELInput", "LNCell", LNCellList,selected = NULL, multiple = TRUE),
selectizeInput("SectorInput", "Sector", SectorList,selected = NULL, multiple = TRUE),
mainPanel(
#plotOutput("distPlot")
verbatimTextOutput("SubRegionText"),
verbatimTextOutput("LCCText"),
verbatimTextOutput("view")
)
)
)
server <- function(input, output) {
observe({
input$SubRegionInput
temp <- SiteInfo[SiteInfo$SITE_SUB_REGION %in% input$SubRegionInput, ]
thisLCCList = sort(temp$BACKHAUL_LCC[!is.na(temp$BACKHAUL_LCC)])
updateSelectizeInput(session = getDefaultReactiveDomain()
, inputId = "LCCInput"
, choices = thisLCCList
, selected= NULL)
thisENBIDList = sort(temp$ENODEB_ID[!is.na(temp$ENODEB_ID)])
updateSelectizeInput(session = getDefaultReactiveDomain()
, inputId = "ENBIDInput"
, choices = thisENBIDList
, selected= NULL)
thisSiteNumberList = sort(temp$SITE_NUMBER[!is.na(temp$SITE_NUMBER)])
updateSelectizeInput(session = getDefaultReactiveDomain()
, inputId = "SiteNumInput"
, choices = thisSiteNumberList
, selected= NULL)
thisSiteNameList = sort(temp$SITE_NAME[!is.na(temp$SITE_NAME)])
updateSelectizeInput(session = getDefaultReactiveDomain()
, inputId = "SiteNameInput"
, choices = thisSiteNameList
, selected= NULL)
thisLNCellList = sort(temp$SECTOR_NUMBER[!is.na(temp$SECTOR_NUMBER)])
updateSelectizeInput(session = getDefaultReactiveDomain()
, inputId = "LNCELInput"
, choices = thisLNCellList
, selected= NULL)
thisSectorList = sort(temp$Sector[!is.na(temp$Sector)])
updateSelectizeInput(session = getDefaultReactiveDomain()
, inputId = "SectorInput"
, choices = thisSectorList
, selected= NULL)
output$view<- renderPrint(temp)
})
Since I do not have access to your data, I used mtcars as an example.
To begin with, since you have so many filtering, I would suggest creating a search or update button, which is what I did in my codes. I only did one filtering using dplyr after extracting all the selectizeInputs. I have to manually change all the empty searching parameter to select all in order to avoid filtering to NA.
Overall, I think the problem with your code was you are observing too many updateSelectizeInputs at once. I did try to recreate using your way, and what I ended with was that I could only update single selectizeInput, and the other selectizeInputs were not selectable.
Hopefully, this method fits your data.
Codes:
library(shiny)
library(dplyr)
library(DT)
data <- mtcars
SubRegionList <- unique(data$cyl)
LCCList <- unique(data$gear)
ENBIDList <- unique(data$am)
SiteNumberList <- unique(data$vs)
# Define UI
ui <- fluidPage(
# Application title
titlePanel("KPI DrillDown"),
# Sidebar with a slider input for number of bins
fluidRow(
selectizeInput("SubRegionInput", "SubRegion/cyl", SubRegionList ,selected = NULL, multiple = TRUE),
uiOutput("LCCInput"),
uiOutput("ENBIDInput"),
uiOutput("SiteNumInput"),
uiOutput("Search"),
mainPanel(
verbatimTextOutput("view")
)
)
)
# Define server logic required
server <- function(input, output, session) {
SiteInfo <- data
# temp <- ""
observe({
if (!is.null(input$SubRegionInput)){
subRegionSelected <- input$SubRegionInput
## Create a temp dataset with the selected sub regions.
temp <- SiteInfo[SiteInfo$cyl %in% subRegionSelected, ]
## Push the newly created selectizeInput to UI
output$LCCInput <- renderUI({
selectizeInput("LCCInput", "LCC/gear", unique(temp$gear), selected = NULL, multiple = TRUE)
})
output$ENBIDInput <- renderUI({
selectizeInput("ENBIDInput", "ENBID/am", unique(temp$am),selected = NULL, multiple = TRUE)
})
output$SiteNumInput <- renderUI({
selectizeInput("SiteNumInput", "SiteNumber/vs", unique(temp$vs), selected = NULL, multiple = TRUE)
})
output$Search <- renderUI({
actionButton("Search", "Search")
})
## Function that linked to the actionButton
display <- eventReactive(input$Search,{
temp <- SiteInfo[SiteInfo$cyl %in% input$SubRegionInput, ]
# ## manually change all the empty searching parameter to select all in order to avoid filtering to NA
LCC <- input$LCCInput
if (is.null(input$LCCInput)){LCC <- unique(temp$gear)}
ENBID <- input$ENBIDInput
if (is.null(input$ENBIDInput)){EBVID <- unique(temp$am)}
SiteNum <- input$SiteNumInput
if (is.null(input$SiteNumInput)){LCC <- unique(temp$vs)}
## Dplyr::filter data
temp <- temp %>%
filter(gear %in% LCC & am %in% ENBID & vs %in% SiteNum)
temp
})
## Run the actionButton
output$view <- renderPrint({
display()
})
} else {
## Display waht the data looks like when no Sub Region is selected
output$view<- renderPrint(data)
}
})
}
# Run the application
shinyApp(ui = ui, server = server)
Related
I am trying to build an interactive data table that changes the displayed columns based on filters chosen by the user. The aim is to have a user select the columns they want to see via a dropdown, which will then cause the datatable to display those columns only.
library(shinyWidgets)
library(DT)
ui <-
fluidPage(
fluidRow(
box(width = 4,
pickerInput(inputId = "index_picker",
label = "Select index/indices",
choices = c("RPI", "RPIX", "CPI", "GDP Deflator"),
selected = "RPI",
multiple = T
)
)
)
fluidRow(
box(DT::dataTableOutput("index_table"), title = "Historic Inflation Indices", width = 12,
solidHeader = T, status = "primary")
)
)
server <- function(input, output, session) {
df_filt <- reactive({
if({
input$index_picker == "RPI" &
!is.null()
})
df_index %>%
select(Period, RPI.YOY, RPI.INDEX)
else if({
input$index_picker == "RPIX"
})
df_index %>%
select(Period, RPIX.YOY, RPIX.INDEX)
})
output$index_table <- renderDataTable({
DT::datatable(df_filt(),
options =
list(dom = "itB",
fixedHeader = T
),
rownames = F
)
})
}
I have similar code to the above that filters based on the row instead, and this works just fine, however, for this column filtering I am getting this error:
Warning in if ({ : the condition has length > 1 and only the first element will be used
I understand that I'm passing a vector to the if statement, but not sure how to recode - would anyone be able to help?
I am trying to generate a shiny app that will first allow the user to (using the notion of dplyr verbs) select the variables they are interested in and then filter those variables based on subsequent selections. I am trying to do this using conditionalPanel() but I am getting stuck finding a way to access the input$ from each conditional panel.
Here is an example:
library('shiny')
library('tidyverse')
library('shinyWidgets')
#Create the data
data <- select(mtcars, c(gear, carb))
#Create page with sidebarlayout
ui <- fluidPage(
sidebarLayout(
sidebarPanel(
#Create picker input where relevant variables are selected
pickerInput(
inputId = 'vars',
label = 'Variables',
choices = colnames(data),
selected = colnames(data),
multiple = T,
pickerOptions(actionsBox = TRUE)
),
#Create conditional panels which show when the variable above is selected
#These panels will be used to filter the data that is selected based on the above variables
conditionalPanel(condition = "input.vars.includes('gear')",
pickerInput(inputId = 'gear',
label = 'Gear',
choices = unique(data$gear),
selected = unique(data$gear),
multiple = T,
pickerOptions(actionsBox = TRUE)
)
),
conditionalPanel(condition = "input.vars.includes('carb')",
pickerInput(inputId = 'carb',
label = 'Carb',
choices = unique(data$carb),
selected = unique(data$carb),
multiple = T,
pickerOptions(actionsBox = TRUE)
)
)
),
mainPanel(
#Show the selected data
verbatimTextOutput('term_selected'),
#Show the selected and filtered data - this won't show
verbatimTextOutput('term_selected_filtered'),
#Try debug with just getting the
verbatimTextOutput('debug_print')
)
)
)
server <- function(input, output) {
#Create the reactive selected data
selected_data <- reactive ({
data %>%
select(input$vars)
})
#Render the selected data
output$term_selected <- renderPrint(selected_data())
#This is where i am stuck
#I need to find a way to access the inputs related to the conditional functions
# selected_filtered_data <- reactive ({
# for (i in length(input$vars)) {
# selected_data() %>%
# filter(input$[first condiitonal panel select] %in% as.symbol(input$vars[i])
# }
# })
#
output$term_selected_filtered <- renderPrint(selected_filtered_data())
#Try to render input input$[first item of input.vars]
output$debug_print <- renderPrint(input$as.symbol(input$vars[1]))
}
shinyApp(ui = ui, server = server)
The problem lies in the server. I have tried input$as.symbol(input$vars[1]) to access the input$gear (assuming that was selected), but it just throws the error: attempt to apply non-function. I tried adding !! as syntactic sugar in front of as.symbol(), but that makes no difference.
I also tried this, in the hope that i could conditionally filter, and had no luck.
selected_filtered_data <- reactive({
selected_data() %>%
if('gear' %in% input$vars) {
filter(gear %in% input$gear) %>%
}
if('carb' %in% input$vars) {
filter(carb %in% input$carb)
}
})
How should I go about doing this?
We may use across (if we want to filter the rows when both column conditions are TRUE) or replace across with if_any (if either one of them is TRUE when they are both selected)
selected_data() %>%
filter(across(all_of(intersect(input$vars,
c('gear', "carb"))), ~ .x %in% input[[cur_column()]]))
-full code
library('shiny')
library('dplyr')
library(tidyr)
library('shinyWidgets')
#Create the data
data <- select(mtcars, c(gear, carb))
#Create page with sidebarlayout
ui <- fluidPage(
sidebarLayout(
sidebarPanel(
#Create picker input where relevant variables are selected
pickerInput(
inputId = 'vars',
label = 'Variables',
choices = colnames(data),
selected = colnames(data),
multiple = TRUE,
pickerOptions(actionsBox = TRUE)
),
#Create conditional panels which show when the variable above is selected
#These panels will be used to filter the data that is selected based on the above variables
conditionalPanel(condition = "input.vars.includes('gear')",
pickerInput(inputId = 'gear',
label = 'Gear',
choices = unique(data$gear),
selected = unique(data$gear),
multiple = T,
pickerOptions(actionsBox = TRUE)
)
),
conditionalPanel(condition = "input.vars.includes('carb')",
pickerInput(inputId = 'carb',
label = 'Carb',
choices = unique(data$carb),
selected = unique(data$carb),
multiple = TRUE,
pickerOptions(actionsBox = TRUE)
)
)
),
mainPanel(
#Show the selected data
verbatimTextOutput('term_selected'),
#Show the selected and filtered data - this won't show
verbatimTextOutput('term_selected_filtered'),
#Try debug with just getting the
verbatimTextOutput('debug_print')
)
)
)
server <- function(input, output) {
#Create the reactive selected data
selected_data <- reactive ({
req(input$vars)
data %>%
select(input$vars)
})
#Render the selected data
output$term_selected <- renderPrint(selected_data())
#This is where i am stuck
#I need to find a way to access the inputs related to the conditional functions
selected_filtered_data <- reactive ({
selected_data() %>%
filter(across(all_of(intersect(input$vars, c('gear', "carb"))), ~ .x %in% input[[cur_column()]]))
})
#
output$term_selected_filtered <- renderPrint(
selected_filtered_data()
)
output$debug_print <- renderPrint(input[[input$vars[1]]])
}
shinyApp(ui = ui, server = server)
-output
I am trying to use a basin, and then update the possible choices of sub-basins within that basin.
However, my code is not working. I cannot make it work neither with observe, nor with reactive, nor with observeEvent nor without all of them.
My ui side is as:
selectInput(inputId = 'countyType_id',
label = '1. Select a basin',
choices = all_basins
),
selectizeInput(inputId = 'subbasins_id',
label = '2. Select subbasins',
choices = subbasins,
selected = head(subbasins, 1),
multiple = TRUE)
and the server side looks like :
observe({
#
# from
# https://shiny.rstudio.com/reference/shiny/latest/updateSelectInput.html
#
subbasins <- sort(unique(curr_spatial$subbasin))
# Can also set the label and select items
updateSelectizeInput(session,
server = FALSE,
"subbasins_id",
label = "2. Select subbasins",
choices = subbasins,
selected = head(subbasins, 1)
)
# It seems the followin has no effect:
# and when it is outside observe, it produces errors!
curr_spatial <- curr_spatial %>%
filter(subbasin %in% input$subbasins_id) %>%
data.table()
})
Any input? please.
I did put the data and the whole code in google drive:
https://drive.google.com/file/d/1qaZG6-VmBhIgMsxs5dffX9PmagkMhuB8/view?usp=sharing
The second selectInput should render from the server and not from the UI to be interactive.
ui.R
selectInput(inputId = 'countyType_id',
label = '1. Select a basin',
choices = all_basins
),
uiOutput('subbasins_id')
server.R
output$subbasins_id <- renderUI({
## add some code to filter subbasin based on the selected basin, i.e. input$countyType_id
curr_spatial <- curr_spatial %>%
filter(subbasin %in% input$subbasins_id) %>%
data.table()
subbasins <- sort(unique(curr_spatial$subbasin))
selectizeInput(inputId = 'subbasins_id',
label = '2. Select subbasins',
choices = subbasins,
selected = head(subbasins, 1),
multiple = TRUE)
})
I am trying to conditionally display filters in my sidebarPanel() if they are selected in an inputSelect() where multiple = TRUE.
Using the example below, the Shiny App should start with neither the "lob_choice" or "segment_choice" filters being displayed in the sidebarPanel, and I have the following requirements:
When both "lob_choice" and "segment_choice" filters are selected in the "filter_choice" filter, both should be displayed.
The "lob_choice" and "segment_choice" filters should only be displayed if they are selected in the "filter_choice" filter.
When either are de-selected, they should be removed from the sidebarPanel display.
It should not matter what order they are selected in in the "filter_choice" filter.
The code below will display one of the conditional panels if it is selected in the "filter_choice" filter, but it won't display either if both are selected at the same time.
ui.R
ui <- fluidPage(
titlePanel("Test App"),
sidebarLayout(
sidebarPanel(
h3("Parameters"),
selectInput("filter_choice",
"In which ways would you like to filter the data?",
c("LOB",
"Segment"),
selected = NULL,
multiple = TRUE),
conditionalPanel(condition = "input.filter_choice == 'LOB'",
selectInput("lob_choice",
"Choose Line(s) of Business:",
c("Brandon", "Kyler", "Trent"),
selected = NULL,
multiple = TRUE)),
conditionalPanel(condition = "input.filter_choice == 'Segment'",
selectInput("segment_choice",
"Choose Segment(s):",
c("LA", "Inverness", "Orlando"),
selected = NULL,
multiple = TRUE))),
mainPanel(tableOutput("table"))
)
)
server.R
library(shiny)
library(dplyr)
server <- function(input, output) {
datasetInput <- reactive({
data <- data.frame(lob = sample(c("Brandon", "Kyler", "Trent"), 1000, replace = TRUE),
segment = sample(c("LA", "Inverness", "Orlando"), 1000, replace = TRUE),
amount = sample(c(100:10000), 1000, replace = TRUE))
# Filter for LOBs if specific LOBs are selected
if (length(input$lob_choice) > 0) {
lob_values <- input$lob_choice
data <- data %>% filter(lob %in% lob_values)
}
# Filter for Segments if specific Segments are selected
if (length(input$segment_choice) > 0) {
segment_values <- input$segment_choice
data <- data %>% filter(segment %in% segment_values)
}
data2 <- aggregate(amount ~ lob + segment, data = data, FUN = sum)
data2[order(-data2$amount), ]
})
output$table <- renderTable({
datasetInput()
})
}
From ?Conditional Panel
condition A JavaScript expression that will be evaluated repeatedly to determine whether the panel should be displayed
So here we can use JS indexOf to check filter_choice
condition = "input.filter_choice !== null && input.filter_choice.indexOf('LOB') >= 0"
#and
condition = "input.filter_choice !== null && input.filter_choice.indexOf('Segment') >= 0"
I'm trying to add a dynamic ggvis plot to a Shiny app. First, user picks a dimension, and then adds items from that dimension.
For global.R and sample data, see https://gist.github.com/tts/a41c8581b9d77f131b31
server.R:
shinyServer(function(input, output, session) {
# Render a selectize drop-down selection box
output$items <- renderUI({
selectizeInput(
inputId = 'items',
label = 'Select max 4. Click to delete',
multiple = TRUE,
choices = aalto_all[ ,names(aalto_all) %in% input$dim],
options = list(maxItems = 4, placeholder = 'Start typing')
)
})
selected <- reactive({
if (is.null(input$items)) {
return(aalto_all)
}
df <- aalto_all[aalto_all[[input$dim]] %in% input$items, ]
df$keys <-seq(1, nrow(df))
df
})
selected %>%
ggvis(~WoS, ~NrOfAuthors, fill = ~School, key := ~keys) %>%
layer_points() %>%
add_tooltip(show_title) %>%
bind_shiny("gv")
show_title <- function(x=NULL) {
if(is.null(x)) return(NULL)
key <- x["keys"][[1]]
selected()$Title20[key]
}
})
ui.R:
shinyUI(fluidPage(
titlePanel('Some (alt)metric data for articles published since 2010'),
sidebarLayout(
sidebarPanel(
selectInput(
inputId = "dim",
label = "Dimension",
choices = dimensions,
selected = c("Title")),
uiOutput("items")
),
mainPanel(
tabsetPanel(
# I'll add more tabs
tabPanel("Plot with ggvis", ggvisOutput("gv"))
)
)
)
))
This is OK
in the beginning, when there are no items selected, and all data is plotted. This is a hack because the ggvis object throws an error if there is no data served.
when all selected items are deleted (which is the same as 1.) and another dimension is chosen
But when I try to switch to another dimension without deleting the items first, I get this:
Error in `$<-.data.frame`(`*tmp*`, "keys", value = c(1L, 0L)) :
replacement has 2 rows, data has 0
I understand that ggvis is very new and constantly developing, but I suspect that there is merely something in Shiny reactive values that is out of sync. If anyone could point out what I'm doing wrong, thanks a lot!
The error is caused because you have a data.frame with zero rows and have a resulting 1:0.
You can change your selected function to:
selected <- reactive({
if (is.null(input$items)) {
return(aalto_all)
}
df <- aalto_all[aalto_all[[input$dim]] %in% input$items, ]
df$keys <-seq_along(df[,1])
if(nrow(df) == 0){
return(aalto_all)
}
df
})