Dialog boxes do not appear for the updated package

[supratim1121992]

I am using R(3.4.4) along with RStudio (1.1.453). The updated version of the package (1.0.0) does not result in dialog boxes appearing for the lists (created using dlgList()). The previous version of the svDialogs package (0.9-57) however still works fine.

[phgrosjean]

Hello,

Could you give more informations about your system, please? (OS, R packages loaded, …), and also a small reproductive example of your problem? Best,

Philippe Grosjean

On 3 Jul 2018, at 09:29, Supratim Chaudhuri notifications@github.com wrote:

I am using R(3.4.4) along with RStudio (1.1.453). The updated version of the package (1.0.0) does not result in dialog boxes appearing for the lists (created using dlgList()). The previous version of the svDialogs package (0.9-57) however still works fine.

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/SciViews/svDialogs/issues/3, or mute the thread https://github.com/notifications/unsubscribe-auth/ADIncVMYCW-JqKNC7tt2WmZ6r0ADB8rEks5uCx10gaJpZM4VAbgd.

[supratim1121992]

I am using Windows 10. I am using your package to develop a UI based analysis module. Before the package got updated, I had dialog boxes appearing for the lists i had created. However, after the package got updated, the dialog boxes stopped appearing and what I get is something like this printed on the console to make a selection.

Outlier detection - Select one or 0 to cancel

1: Detect outliers using standard deviation (6 sigma) 2: Detect outliers using Interquartile range (IQR)

Moreover, the messages have lost the indentation I had used with the previous version of the package.

If you want, you can even use the package I created and check. The following code should give you an idea of what has been happening.

library(devtools) install_github("supratim1121992/AutoAnalyse") library(AutoAnalyse)

data("Titanic") dt_proc<-Pre_Proc(dataset = Titanic) Auto_EDA(dataset = dt_proc)

I have to remove the current version of svDialogs and install the previous version (0.9-57) to be able to work with it as before.

[phgrosjean]

Support for RStudio was introduced in version 1.0.0 of svDialogs. It works as follows:

• If a graphical dialog box exists, use it,
• If it does not exists, or the context could be problematic, fall back to a textual version at the R console.

This depends on the dialog box. For instance, there are always graphical message boxes, or input boxes available in RStudio, but no built-in list dialog boxes.

So the code checks if it is RStudio Desktop or Server. If it is Desktop, it uses the standard graphical list dialog box. If it is the Server version, there is no guarantee that the graphical dialog box will be available, and if it is, it may appear on the server, not on the client’s computer. So, we fall back to the textual version.

So, unless my code to detect the RStudio version is incorrect, I deduce that you use RStudio Server (is it?), and in this case, svDialogs behaves as expected, see ?dlgList.

Best,

Philippe Grosjean

On 3 Jul 2018, at 10:52, Supratim Chaudhuri notifications@github.com wrote:

I am using Windows 10. I am using your package to develop a UI based analysis module. Before the package got updated, I had dialog boxes appearing for the lists i had created. However, after the package got updated, the dialog boxes stopped appearing and what I get is something like this printed on the console to make a selection. Moreover, the messages have lost the indentation I had used with the previous version of the package.

`Outlier detection - Select one or 0 to cancel

1: Detect outliers using standard deviation (6 sigma) 2: Detect outliers using Interquartile range (IQR)`

If you want, you can even use the package I created and check. The following code should give you an idea of what has been happening.

`library(devtools) install_github("supratim1121992/AutoAnalyse") library(AutoAnalyse)

data("Titanic") dt_proc<-Pre_Proc(dataset = Titanic) Auto_EDA(dataset = dt_proc)`

I have to remove the current version of svDialogs and install the previous version (0.9-57) to be able to work with it as before.

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/SciViews/svDialogs/issues/3#issuecomment-402062890, or mute the thread https://github.com/notifications/unsubscribe-auth/ADIncfHZ4EMgqmYVrx0oy_jCX6WoHzyHks5uCzC6gaJpZM4VAbgd.

[supratim1121992]

I'm using RStudio Desktop and like I said I don't have this issue with the previous version of your package. Only the version 1.0.0 causes this issue. If I use

remove.packages("svDialogs")

and reinstall v0.9-57 (from .tar file), everything works fine.

[phgrosjean]

On 3 Jul 2018, at 12:07, Supratim Chaudhuri notifications@github.com wrote:

I'm using RStudio Desktop and like I said I don't have this issue with the previous version of your package.

OK, I will lok at this, but for now, I am at a conference. Best,

Phiippe Grosjean

Only the version 1.0.0 causes this issue. If I use

remove.packages("svDialogs")

and reinstall v0.9-57 (from .tar file), everything works fine.

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/SciViews/svDialogs/issues/3#issuecomment-402090515, or mute the thread https://github.com/notifications/unsubscribe-auth/ADIncYZulqsTqsTi5MnmBKJBE2nfn732ks5uC0JxgaJpZM4VAbgd.

[supratim1121992]

Hi Philippe

Did you get a chance to inspect the issue?

[phgrosjean]

Could you try with the latest devel version? I think it solves your problem. In case you don't want to compile from source, you can get Windows binaries here: https://ci.appveyor.com/api/buildjobs/55iqs8g95xmhk9ms/artifacts/svDialogs_1.0.1.zip

[supratim1121992]

Thanks Philippe for getting back to me. The dlgList function works as expected with version 1.0.1. However, the messages don't show proper indentation. For instance, if I want two messages printed in separate lines I'm using dlgMessage(message = c("message1","message2"),type = "ok" But that does not introduce a line break. Should I use '\n' instead?

[phgrosjean]

That is another issue. I can confirm this under RStudio. However, the dlgMessage() code includes '\n' as expected, but the RStudio message box does not honor it. It is a limitation, apparently, with RStudio message box, and can do nothing in svDialogs to solve this. This is now specified in ?dlgMessage Since the initial problem seems to be solved, I close the issue.

[supratim1121992]

OK thank you for that.