Shiny app - sending ompr console prints to Shiny app

[datadrivensupplychain]

Maybe this isn't the right forum but SO has failed me. I'm developing Shiny apps that perform supply chain network optimization "under the hood" using ompr and SYMPHONY. I personally like to set the solver to verbose, so I can track solve progress on the console.

Has anyone used ompr with a Shiny app and been able to print the console content to the Shiny app "live"? That would be ideal.

Thanks
Ralph

[dirkschumacher]

Hey Ralph,

this should give you a good starting point (if I understood your problem correctly). You can use capture.output to capture the logs.

library(shiny)
library(ompr)
library(ompr.roi)
library(ROI.plugin.glpk)

ui <- bootstrapPage(
  numericInput("n", "Number of items", 10),
  verbatimTextOutput("text")
)

server <- function(input, output) {
  output$text <- renderText({
    req(input$n >= 1)
    text <- capture.output({
      MIPModel() |> 
        add_variable(x[i], i = 1:input$n, type = "binary") |> 
        set_objective(sum_over(runif(1) * x[i], i = 1:input$n), "max") |> 
        add_constraint(sum_over(runif(1) * x[i], i = 1:input$n) <= 2) |> 
        solve_model(with_ROI("glpk", verbose = TRUE))
    })
    paste0(text, collapse = "\n")
  })
}

shinyApp(ui = ui, server = server)

[dirkschumacher]

However, this might not give you a live view into the logs. 🤔

[dirkschumacher]

Not sure how, but I would look into capture.output and see if it is possible to live capture the output and append it to a file as soon as it changes. Then use reactiveFileReader to watch that file. You might need to start the solver in a background process using callr or shiny's async model.

[datadrivensupplychain]

Thanks Dirk. Yes I want to track the console output live to understand the solver progress (doesn't really matter if I get the console output after the solve completes.). I look more into async in a couple weeks when I have some more time for R&D. Right now I have my code set up to save the results of solve_model as an object in the workspace, then I use get_solution to extract the decision variables into a few different data frames. I then do postprocessing on those data frames for the Shiny output. I just did a webinar with RStudio "R in Supply Chain Management" that demos the app. I'll post a link when the recording is on YouTube. Ralph Asher ***@***.***

…

On Thu, Jan 20, 2022, 2:50 AM Dirk Schumacher ***@***.***> wrote: Not sure how, but I would look into capture.output and see if it is possible to live capture the output and append it to a file as soon as it changes. Then use reactiveFileReader to watch that file. You might need to start the solver in a background process using callr or shiny's async model. — Reply to this email directly, view it on GitHub <#387 (reply in thread)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ASKSE664PUOC6N3QE76MCJDUW7EEDANCNFSM5MLFJ5DQ> . Triage notifications on the go with GitHub Mobile for iOS <https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675> or Android <https://play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub>. You are receiving this because you authored the thread.Message ID: ***@***.***>