added code for plotly

[winsonfzyang]

Hello! I think I might know a way to allow plotly to work on your app.
you would have to add this to server.R:

output$plotly.plot <- renderPlotly({ ggplotly(plotObject()) })

output$ui_plotly <- renderUI({ plotlyOutput('plotly.plot', width = "100%" ,height = input$height) })

But doing that your box plots will not work because when you have no input in Groups, your group is an empty column, and plotly cannot read it.

geom_boxplot(aes(group=NULL)

My suggestion to get it working is to replace NULL with input$x:

geom_boxplot(aes_string(group=input$x)

[smouksassi]

Thanks.Will have a look. But there is use cases where you don't want the group to be x for example you might want x and the factor color you chose. Ggplot automatically choose the combination of factors and I have this to be able to have boxplots and overlay a median lines colored by a factor.Plotly is in part implemented but indeed having more things working is desirable. From: notifications@github.comSent: March 23, 2018 7:27 AMTo: ggplotwithyourdata@noreply.github.comReply-to: reply@reply.github.comCc: subscribed@noreply.github.comSubject: [smouksassi/ggplotwithyourdata] added code for plotly (#19) Hello! I think I might know a way to allow plotly to work on your app. you would have to add this to server.R: output$plotly.plot <- renderPlotly({ ggplotly(plotObject()) }) output$ui_plotly <- renderUI({ plotlyOutput('plotly.plot', width = "100%" ,height = input$height) }) But doing that your box plots will not work because when you have no input in Groups, your group is an empty column, and plotly cannot read it. geom_boxplot(aes(group=NULL) My suggestion to get it working is to replace NULL withinput$x: geom_boxplot(aes_string(group=input$x) —You are receiving this because you are subscribed to this thread.Reply to this email directly, view it on GitHub, or mute the thread.

[smouksassi]

Dear winsonfzyang ,
Can you show some example plotly outputs you would like the ggquickeda ( new package name to produce ?) will repoen in the new repo thanks

[winsonfzyang]

Hello smouksassi,
Maybe a scatterplot or box plot for a start will be good. I've made my own modifications to your app for my usage, and I think this will work

[smouksassi]

Dear winsonfzyang, please comment on the issue with ggquickeda the new version of this package just submitted to CRAN. Plotly will work whenever you remove the ignore mapped grouping and you map a group explicilty but then you lose the automatic grouping that ggplot do under the hood.

Using the built in sample_data I am using Gender as x variable and conc as y variable and Gender as Colour B, Group By and Fill by: and we split column by Race

now suppose you want another plot

when grouping is not specified ggplot knows that we wnat to group by the combination of x and color

this where Plotly will not work unless you do the Race_gender group and group by it as below
you will get the plotly but it will not be dodged probalby because it is automatic in ggplot

This illustrate why just grouping by x variable might not be enough and this app was designed with ggplot first and foremost in mind. I would like to make it more plotly compatible by writing directly in plotly as ggplotly has some limitations.