Myujth

I am trying to build a Shiny App that will do the following:

1. Read a time-series data set from a database (e.g. DateTime, Value, Status), where Status flags whether a point is valid and be plotted


2. Create a time-series plot of that data (with only Status = Valid points showing).


3. Create a data table in the interface showing the data that are being plotted


4. Allow for the user to review the data in the plot and then "flag out" points by setting the Status in the table to invalid

Steps 1-3 above I understand how to do, but I'm looking for suggestions on how to implement the 4th step. Can Shiny have an interactive table where if a user sets the stats flag, it will react and redraw the plot?

The idea here is to make an interactive data cleaning tool for time-series data recorded from an instrument that often will have bad data points in it.

Thanks

Here is a solution using library(DT).
Type '0' in the 'plot'-column to remove the datapoint from the plot.

library(shiny)
library(DT)

shinyApp(
 ui = fluidPage(DTOutput('myTable'),
 plotOutput("myPlot")),
 server = function(input, output) 
 output$tbl = renderDT(iris,
 options = list(lengthChange = FALSE),
 editable = TRUE)

 myData = data.frame(y = 1:10, plot = 1)

 output$myTable = renderDT(myData, selection = 'none', editable = TRUE)

 proxy = dataTableProxy('myTable')

 observeEvent(input$myTable_cell_edit, 
 info = input$myTable_cell_edit
 str(info)
 i = info$row
 j = info$col
 v = info$value
 myData[i, j] <<- DT::coerceValue(v, myData[i, j])
 replaceData(proxy, myData, resetPaging = FALSE) # important
 )

 output$myPlot <- renderPlot(
 input$myTable_cell_edit
 plot(myData$y[myData$plot > 0], myData$y[myData$plot > 0])
 )

 
)

For further information see this.