flexdashboard: shared processing among reactive output chunks

[John Williams]

I have common, time-consumming code (with reactive input)  that needs to run (within reactive output objects)  in multiple chunks in a flexdashboard 

How can I:

• process the code once, sharing results among multiple reactive chunks?,  and/or
• express the code as a function, to be called from each reactive object?,  and/or
• other way(s)?

Thanks!

[John Williams]

Maybe with a reactive observer?  I.e., in a flex dashboard, create a reactive obsserver in one code chunck, and access that reactive observer in subsequent code chunk, from within a reactive output function?

[Joe Cheng]

Close--not a reactive observer (observe() and observeEvent()), but a reactive expression (reactive()).

--
You received this message because you are subscribed to the Google Groups "Shiny - Web Framework for R" group.
To unsubscribe from this group and stop receiving emails from it, send an email to shiny-discus...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/shiny-discuss/0bd81e11-b197-4647-a5e7-3acb432ad023%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[John Williams]

got it -  had to wrap code in a function before wrapping in reactive() - thank you,  Joe Chang & Rstudio,  for Shiny - its awesome!!!


On Monday, June 20, 2016 at 10:39:44 PM UTC-4, John Williams wrote:

CORRECTION:

Keep getting:  "Error: data must be a data frame class object"  ... caused by the renderSimpleNetwork() in 2nd chunk ...

```{r  flexdashboardChunk1}

reactiveCode <- quote(

{

 code producing a dataframe (using reactive input)  

 }

)    

```

```{r flexdashboardChunk2}

values <- reactiveValues()

values.topdocs <- reactiveCode

renderSimpleNetwork({

  simpleNetwork(values.topdocs, Source = "src", Target = "tar", linkDistance = 150, fontSize = 12)

  })

````

On Monday, June 20, 2016 at 10:36:34 PM UTC-4, John Williams wrote:

Keep getting:  "Error: data must be a data frame class object"  ... caused by the 

```{r  flexdashboardChunk1}

reactiveCode <- quote(

{

 code producing a dataframe (using reactive input)  

 }

)    

```

```{r flexdashboardChunk2}

values <- reactiveValues()

values.topdocs <- topDocsandProcs

renderSimpleNetwork({

  simpleNetwork(values.topdocs, Source = "src", Target = "tar", linkDistance = 150, fontSize = 12)

  })

````

On Monday, June 20, 2016 at 6:18:35 PM UTC-4, John Williams wrote:

[John Williams]

well, not so fast ... it seems that my reactive function - reactiveTDPS() - is fully executing each time it is called, not just when inputs change --- anybody know how to correct this? 

also,  how/where  to implement progress indicator after initial loading?

app and code here   

Thanks!

[Joe Cheng [RStudio]]

I'm not sure where the confusion is without a more complete code sample, but, there's not much different here vs. any other server function (just imagine the entire R Markdown document is a single Shiny server function).

reactiveTDPS <- reactive({

  ... yield a data frame...

})

renderSimpleNetwork({

  simpleNetwork(reactiveTDPS(), ...)

})

It doesn't matter if these two pieces of code are in the same or different code chunks.