My current workflow in a shiny application is to run a R script as a cron job periodically to pull various tables from multiple databases as well as download data from some APIs. These are then saved as a .Rdata file in a folder called data.
In my global.R file I load the data by using load("data/workingdata.Rdata"). This results in all the dataframes (about 30) loading into the environment. I know I can use the reactiveFileReader() function to refresh the data, but obviously it would have to be used in the server.R file because of an associated session with the function. Also, I am not sure if load is accepted as a readFunc in reactiveFileReader(). What should be the best strategy for the scenario here?
This example uses a reactiveVal object with observe and invalidateLater. The data is loaded into a new environment and assigned to the reactiveVal every 2 seconds.
library(shiny)
ui <- fluidPage(
actionButton("generate", "Click to generate an Rdata file"),
tableOutput("table")
)
server <- shinyServer(function(input, output, session) {
## Use reactiveVal with observe/invalidateLater to load Rdata
data <- reactiveVal(value = NULL)
observe({
invalidateLater(2000, session)
n <- new.env()
print("load data")
env <- load("workingdata.Rdata", envir = n)
data(n[[names(n)]])
})
## Click the button to generate a new random data frame and write to file
observeEvent(input$generate, {
sample_dataframe <- iris[sample(1:nrow(iris), 10, F),]
save(sample_dataframe, file="workingdata.Rdata")
rm(sample_dataframe)
})
## Table output
output$table <- renderTable({
req(data())
data()
})
})
shinyApp(ui = ui, server = server)
A few thoughts on your workflow:
In the end with your RData-approach you are setting up another data source in parallel to your databases / APIs.
When working with files there always is some housekeeping-overhead (e.g. is your .RData file completed when reading it?). In my eyes this (partly) is what DBMS are made for – taking care about the housekeeping. Most of them have sophisticated solutions to ensure that you get what you query very fast; so why reinvent the wheel?
Instead of continuously creating your .RData files and polling data with the reactiveFileReader() function you could directly query the DB for changes using reactivePoll (see this
for an example using sqlite). If your queries are long running (which I guess is the cause for your workflow) you can wrap them in a future and run them asynchronously (see this post
to get some inspiration).
Alternatively many DBMS provide something like materialized views to avoid long waiting times (according user privileges presumed).
Of course, all of this is based on assumptions, due to the fact, that your eco-system isn’t known to me, but in my experience reducing interfaces means reducing sources of error.
You could use load("data/workingdata.Rdata") at the top of server.R. Then, anytime anyone starts a new Shiny session, the data would be the most recent. The possible downsides are that:
there could be a hiccup if the data is being written at the same time a new Shiny session is loading data.
data will be stale if a session is open just before and then after new data is available.
I imagine the first possible problem wouldn't arise enough to be a problem. The second possible problem is more likely to occur, but unless you are in a super critical situation, I can't see it being a substantial enough problem to worry about.
Does that work for you?
Related
Introduction
I have created an R shiny dashboard app that is quickly getting quite complex. I have over 1300 lines of code all sitting in app.R and it works. I'm using RStudio.
My application has a sidebar and tabs and rather than using modules I dynamically grab the siderbar and tab IDs to generate a unique identifier when plotting graphs etc.
I'm trying to reorganise it to be more manageable and split it into tasks for other programmers but I'm running into errors.
Working Code
My original code has a number of library statements and sets the working directory to the code location.
rm(list = ls())
setwd(dirname(rstudioapi::getActiveDocumentContext()$path))
getwd()
I then have a range of functions that sit outside the ui/server functions so are only loaded once (not reactive). These are all called from within the server by setting the reactive values and calling the functions from within something like a renderPlot. Some of them are nested, so a function in server calls a function just in regular app.R which in turn calls another one. Eg.
# Start of month calculation
som <- function(x) {
toReturn <- as.Date(format(x, "%Y-%m-01"))
return(toReturn)
}
start_fc <- function(){
fc_start_date <- som(today())
return(fc_start_date)
}
then in server something like this (code incomplete)
server <- function(input, output, session) {
RV <- reactiveValues()
observe({
RV$selection <- input[[input$sidebar]]
# cat("Selected:",RV$selection,"\r")
})
.......
cat(paste0("modelType: ",input[[paste0(RV$selection,"-modeltype")]]," \n"))
vline1 <- decimal_date(start_pred(input[[paste0(RV$selection,"-modeltype")]],input[[paste0(RV$selection,"-modelrange")]][1]))
vline2 <- decimal_date(start_fc())
.......
Problem Code
So now when I take all my functions and put them into different .R files I get errors indicating the functions haven't been loaded. If I load the source files by highlighting them and Alt-Enter running them so they are loaded into memory then click on Run App the code works. But if I rely on Run App to load those source files, and the functions within them, the functions can't be found.
source('./functionsGeneral.R')
source('./functionsQuote.R')
source('./functionsNewBusiness.R')
source('./ui.R')
source('./server.R')
shinyApp(ui, server)
where ui.R is
source('./header.R')
source('./sidebar.R')
source('./body.R')
source('./functionsUI.R')
ui <- dashboardPage(
header,
sidebar,
body
)
Finally the questions
In what order does R Shiny Dashboard run the code. Why does it fail when I put the exact same inline code into another file and reference it with source('./functions.R')? Does it not load into memory during a shiny app session? What am I missing?
Any help on this would be greatly appreciated.
Thanks,
Travis
Ok I've discovered the easiest way is to create a subfolder called R and to place the preload code into that folder. From shiny version 1.5 all this code in the R folder is loaded first automatically.
I'm quite new to Shiny, so my apologizes if my question is an easy one. I tried to check on google and stackoverflow but couldn't locate a simple and helpful answer so far.
What's my goal/issue: I'm coding a Shiny page that displays a table with hundreds of thousands of rows.
Data is sourced from different databases, manipulated, cleaned, and displayed to all the users upon request.
Problem 1: in order to load all the data, the script takes almost 5minutes
Problem 2: if at 8:00am user1 requests this data and at 8:05am user2 requests the same data, two different queries are launched and also two different spaces in memory are used to show exactly the same data to two different users.
So the question is: shall I use a cache system to enhance this process?
if not, what else shall I use?
I found a lot of official Shiny documentation on caching plots but nothing related to caching data (and I found this quite surprising).
Other useful information: data in cache should be deleted every evening around 10pm since new data will be available the next day / early morning.
Code:
ui <- dashboardPage( # https://rstudio.github.io/shinydashboard/structure.html
title = "Dashboard",
dashboardHeader(title = "Angelo's Board"),
dashboardSidebar( # inside here everything that is displayed on the left hand side
includeCSS("www/styles.css"),
sidebarMenu(
menuItem('menu 1', tabName = "menu1", icon = icon("th"),
menuItem('Data 1', tabName = 'tab_data1'))
)),
dashboardBody(
tabItems(
tabItem(tabName = 'tab_data1')),
h3("Page with big table"),
fluidRow(dataTableOutput("main_table"))
))
server <- function(input, output, session) {
output$main_tabl <- renderDataTable({
df <- data.frame(names = c("Mark","George","Mary"), age = c(30,40,35))
})
}
cat("\nLaunching 'shinyApp' ....")
shinyApp(ui, server)
Resources I used to check for potential solution:
How to cache data in shiny server? but apparently I cannot use Jason Bryer package
https://shiny.rstudio.com/reference/shiny/1.2.0/memoryCache.html but I have no idea of how to use this code applied to my example
https://shiny.rstudio.com/articles/plot-caching.html is mainly focused on plot caching
Any help would be much appreciated. Thanks
I would break out the bulk of your ETL processes into a separate R script and set that script to run on a cron. You can then have this script write out the processed dataframe(s) to a .feather file. Then have your shiny app load the feather file(s) - feather is optimized for reading so should be fast.
Example, take the necessary libraries and code out of your server.R (or app.R) file, and create a new R script called query.R. That script performs all the ETL operations and finally writes out your data to a .feather file (requires the feather package). Then create a crontab to run that script as often as needed.
Your server.R script then just needs to read in that feather file when the app loads and you should see a significant performance improvement. In addition, you have have the query.R script run during off hours so that performance on the linux box isn't negatively impacted.
Another option, put this DataFrame in global.R and change /etc/shiny-server/shiny-server.conf by adding «app_idle_timeout 0» after «location / {». This will disable application idle timeouts in Shiny Server, so global.R will be in RAM for all users.
To prevent first user from long data loading, you can put in cron «#reboot wget -O index.html localhost:3838» on your server, so on every reboot global.R will load to memory automatically.
Also, about pre-cache organisation you can read here.
I'm working on a complex (for me) Shiny app ~2500 lines of code. The basic structure is as follows, though I can't necessarily share a reproducible example since most of the information is confidential.
Right now, I am using df1 <- read.csv() etc to read in several CSV files as dataframes. I want to use reactiveFileReader() to make it such that the dataframes automatically update when the source CSV file is modified. I think my problem may be related to the fact that I am not doing this in a reactive context, but there is a reason for this. I am using the dataframes df1 etc to perform many calculations and to create new variables throughout the app (UI and server sections).
It also might be important to note that I am doing these file imports in the UI part of the Shiny app, since I need to rely on the factor levels of these dataframes to populate drop down selectInput in my UI. This might not be necessary.
Here is what I have tried (although I am pretty lost):
reader <- reactiveFileReader(intervalMillis = 1000, filePath =
"Data_Record.csv", readFunc = read.csv)
data_record <- reactive({
data_df <- reader()
return(data_df)
})
What I was expecting was for data_record to be a dataframe containing the information from the CSV, but it ends up being a "reactive expression". When I try to perform operations on data_record, like subsetting, I receive errors since that variable is not a dataframe.
Is there any way for me to update these dataframes upon modification to the underlying CSV outside of a reactive context? Even a scheduled update like every 10 seconds or so would work as well. My ultimate goal are dataframes that update when a CSV is modified, but scheduled updates are fine as well.
Thanks in advance for all the help and I apologize for not being able to supply a reproducible example! I will try to add more information as needed.
So if you want the data to be reactive, it has to be imported in the server section of the app as a 'reactive'. In shiny, 'reactives' become functions so to do anything with them you have to reference them their name followed by parenthesis inside a reactive function (reactive, observe, render etc).
For example, with your code above, reader becomes a reactive data frame. You can perform normal data manipulation on reader if you follow the rules of reactives outlined above.
# server.R
reader <- reactiveFileReader(intervalMillis = 1000, filePath =
"Data_Record.csv", readFunc = read.csv)
filtered_reader_df <- reactive({
reader() %>% filter(x > 100)
})
Where filtered_reader_df becomes a reactive filtered version of the reactive csv file. Again, to use filtered_reader_df in subsequent reactive function it must be referenced as filtered_reader_df() as it is a reactive function itself.
Finally, you can still use the reactive csv file to populate UI elements with the updateSelectInput() function inside an observer in the server. For example:
ui <- fluidPage(
selectInput("mySelectInput", "Select")
)
server <- function(input, output, session) {
reader <- reactiveFileReader(intervalMillis = 1000, filePath =
"Data_Record.csv", readFunc = read.csv)
observe({
select_input_choices <- unique(reader()$factor_column)
updateSelectInput(session, inputId = "mySelectInput", choices = select_input_choices)
})
}
The code above will update the choices of the select input every time the reader() data frame changes with the reactiveFileReader where unique(reader()$factor_column) are the reactive unique values of the factor column you want to populate the input with.
I am working on a shiny app that will read a few RData files in and show tables with the contents. These files are generated by scripts that eventually turns the data into a data frame. They are then saved using the save() function.
Within the shiny application I have three files:
ui.R, server.R, and global.R
I want the files to be read on an interval so they are updated when the files are updated, thus I am using:
reactiveFileReader()
I have followed a few of the instructions I have found online, but I keep getting an error "Error: missing value where TRUE/FALSE is needed". I have tried to simplify this so I am not using:
reactiveFileReader()
functionality and simply loading the file in the server.R (also tried in the global.R file). Again, the
load()
statement is reading in a data frame. I had this working at one point by loading in the file, then assigning the file to a variable and doing an "as.data.table", but that shouldn't matter, this should read in a data frame format just fine. I think this is a scoping issue, but I am not sure. Any help? My code is at:
http://pastebin.com/V01Uw0se
Thanks so much!
Here is a possible solution inspired by this post http://www.r-bloggers.com/safe-loading-of-rdata-files/. The Rdata file is loaded into a new environment which ensures that it will not have unexpected side effect (overwriting existing variables etc). When you click the button, a new random data frame will be generated and then saved to a file. The reactiveFileReader then read the file into a new environment. Lastly we access the first item in the new environment (assuming that the Rdata file contains only one variable which is a data frame) and print it to a table.
library(shiny)
# This function, borrowed from http://www.r-bloggers.com/safe-loading-of-rdata-files/, load the Rdata into a new environment to avoid side effects
LoadToEnvironment <- function(RData, env=new.env()) {
load(RData, env)
return(env)
}
ui <- shinyUI(fluidPage(
titlePanel("Example"),
sidebarLayout(
sidebarPanel(
actionButton("generate", "Click to generate an Rdata file")
),
mainPanel(
tableOutput("table")
)
)
))
server <- shinyServer(function(input, output, session) {
# Click the button to generate a new random data frame and write to file
observeEvent(input$generate, {
sample_dataframe <- data.frame(a=runif(10), b=rnorm(10))
save(sample_dataframe, file="test.Rdata")
rm(sample_dataframe)
})
output$table <- renderTable({
# Use a reactiveFileReader to read the file on change, and load the content into a new environment
env <- reactiveFileReader(1000, session, "test.Rdata", LoadToEnvironment)
# Access the first item in the new environment, assuming that the Rdata contains only 1 item which is a data frame
env()[[names(env())[1]]]
})
})
shinyApp(ui = ui, server = server)
Ok - I figured out how to do what I need to. For my first issue, I wanted the look and feel of 'renderDataTable', but I wanted to pull in a data frame (renderDataTable / dataTableOutput does not allow this, it must be in a table format). In order to do this, I found a handy usage of ReportingTools (from Bioconductor) and how they do it. This allows you to use a data frame directly and still have the HTML table with the sorts, search, pagination, etc.. The info can be found here:
https://bioconductor.org/packages/release/bioc/html/ReportingTools.html
Now, for my second issue - updating the data and table regularly without restarting the app. This turned out to be simple, it just took me some time to figure it out, being new to Shiny. One thing to point out, to keep this example simple, I used renderTable rather than the solution above with the ReportingTools package. I just wanted to keep this example simple. The first thing I did was wrap all of my server.R code (within the shinyServer() function) in an observe({}). Then I used invalidateLater() to tell it to refresh every 5 seconds. Here is the code:
## server.R ##
library(shiny)
library(shinydashboard)
library(DT)
shinyServer(function(input, output, session) {
observe({
invalidateLater(5000,session)
output$PRI1LastPeriodTable <- renderTable({
prioirtyOneIncidentsLastPeriod <- updateILP()
})
})
})
Now, original for the renderTable() portion, I was just calling the object name of the loaded .Rdata file, but I wanted it to be read each time, so I created a function in my global.R file (this could have been in server.R) to load the file. That code is here:
updateILP <- function() {
load(file = "W:/Projects/R/Scripts/ITPOD/itpod/data/prioirtyOneIncidentsLastPeriod.RData", envir = .GlobalEnv)
return(prioirtyOneIncidentsLastPeriod)
}
That's it, nothing else goes in the global.R file. Your ui.R would be however you have it setup, call tableOutout, dataTableOutput, or whatever your rendering method is in the UI. So, what happens is every 5 seconds the renderTable() code is read every 5 seconds, which in turns invokes the function that actually reads the file. I tested this by making changes to the data file, and the shiny app updated without any interaction from me. Works like a charm.
If this is inelegant or is not efficient, please let me know if it can be improved, this was the most straight-forward way I could figure this out. Thanks to everyone for the help and comments!
I'm working on an app in R where the users need to choose a file from their computer, with a RShiny fileInput button. I want to modify this, so that the associated variable can be assigned (i.e. a file can be loaded) automatically by the programm, without having the user click on the button and choose the file.
The problem I'm facing is that a fileInput has 4 fields, amongst which I only can know 3. For instance, when I load the file hello.csv in the variable inFile through the normal procedure, here is what I get :
inFile$name = hello.csv
inFile$size = 8320
inFile$type = text/csv
inFile$datapath = C:\\Users\\MyName\\AppData\\Local\\Temp\\Rtmpkh8Zcb/7d5f0ff0111d440c7a66b656/0
Though I could have guessed the second and the third one knowing the file, I have no idea how the datapath field is assigned...
I've tried to declare inFile as a NULL global variable, then to assign one by one the different fields, but I'm stuck with this last one. Is there an other way to do, like a function that mimics the behaviour of a user who clicks on the file input button and choose a specified file ?
Thank you very much.
If all you're looking to do is load a file initially, you don't have to rely on Shiny functions to do that. You can just rely on R functions. Set up your app like this:
ui <- shinyUI(
fileInput("inFile", label="Choose a file", multiple=F)
)
server <- shinyServer(function(input, output, session) {
values <- reactiveValues()
dat <- reactive({
if (is.null(inFile$datapath)) {
dat <- read.csv("path/to/your.csv")
values$file_name = "your.csv"
values$file_type = "csv"
values$file_size = file.size("path/to/your.csv")
values$file_path = "path/to/your.csv"
} else {
dat <- read.csv(inFile$datapath)
values$file_name = inFile$name
values$file_size = inFile$size
values$file_type = inFile$type
values$file_path = inFile$datapath
}
})
})
shinyApp(ui=ui, server=server)
In the above code, the Shiny app will start and see that inFile$datapath is NULL and will load a predefined file of your choosing. It won't run again until inFile changes, at which point it will load the file that the user pointed to.
Hope that helps.
Update
I changed the code above to use reactiveValues to store the pieces of information that need to be used throughout the app. If you just set those and then do a find/replace for input$inFile$datapath and replace it values$file_path, your code should work just fine.
Here is how I figured it out :
I edited the original code, so that all the read.csv(...) are replaced with calls to a data.frame global variable. I also added a small button that you need to click on before you continue. This button saves what you just loaded in the Database (if you chose a file with the fileInput) and assigns the right values to the global variables that will be needed for the following operations. If you chose no file at all, it will directly assign the variables from the data found in the Database.
So I did not find a proper solution to the problem, but this is a workaround that will do the job in my case.
#brittenb I couldn't get your reactive solution to work as I wanted to, that's why I ended up doing this another way. Thanks for having taken the time to think about it though.
I'm still open to suggestions on how to update the file in a fileInput without user interaction.