[Bioc-devel] 80% of man pages must have runnable examples (package with numerous Shiny apps)

Hi Lori and Kevin:

Thank you for your helpful responses! I was able to use "if
(interactive()) { ... }" format and return the appDir object for the
user. Thanks again!

Lindsay
On Mon, Nov 19, 2018 at 3:02 PM Kevin RUE <kevinrue67 at gmail.com> wrote:

Hi Lindsay,

Check out https://github.com/csoneson/iSEE/blob/master/R/iSEE-main.R

function "iSEE" for an example of both the function return value, and the
man page.

First, I would suggest that your function _returns_ the "appDir" object

to the user, and that you leave it to the user to call "shiny::runApp" with
options appropriate to their system or preferences (e.g
launch.browser=FALSE, port=1234, etc).

Second, for the man page, you don't have to put the _entire_ @example

block inside \dontrun{}. You can put everything that doesn't launch the
Shiny app outside the \dontrun{} block and only put the one "shiny" line
inside the \dontrun{}. The one line will represent less than 80% of the man
page.

However, even better, you can avoid the \dontrun{} option altogether and

put the "shiny" statement within a "if (interactive()) { ... }" block.

So adapting you existing code, I would have
#' @examples
#' # Example 1: Create an interactive litre plot for the logged data

using

#' # default background of hexagons.
#'
#' data(soybean_ir_sub)
#' data(soybean_ir_sub_metrics)
#' soybean_ir_sub_log <- soybean_ir_sub
#' soybean_ir_sub_log[,-1] <- log(soybean_ir_sub[,-1]+1)
#' if (interactive()){
#' plotLitreApp(data = soybean_ir_sub_log, dataMetrics =

soybean_ir_sub_metrics)

#' }
#'
#' # Example 2: Repeat the same process, only now plot background data as
#' # individual points. Note this may be too slow now that all points

are drawn

#' # in the background.
#' if (interactive()){
#' plotLitreApp(data = soybean_ir_sub_log, dataMetrics =

soybean_ir_sub_metrics,

#' option = "allPoints", pointColor = "red")
#' }

However, once more, I suggest that you return "appDir", and you would

then have something like:

#' app <- plotLitreApp(data = soybean_ir_sub_log, dataMetrics =

soybean_ir_sub_metrics)

#' if (interactive()){
#'    shiny::runApp(app, [options defined by your user])
#' }

I hope this helps. Otherwise, let me know and I can chime in the review

of your package

Best wishes
Kevin

On Mon, Nov 19, 2018 at 7:00 PM Shepherd, Lori <

Lori.Shepherd at roswellpark.org> wrote:

Submit with the dontrun{} and temporarily ignore the ERROR -  when

submitting please reference the explanation below.  Your reviewer could
provide more information and will decide how to proceed -  generally
examples will be run manually to check for accuracy and an exception can be
made.

Lori Shepherd

Bioconductor Core Team

Roswell Park Cancer Institute

Department of Biostatistics & Bioinformatics

Elm & Carlton Streets

Buffalo, New York 14263

Rutter <lindsayannerutter at gmail.com>

Sent: Monday, November 19, 2018 1:49:50 PM
To: bioc-devel at r-project.org
Subject: [Bioc-devel] 80% of man pages must have runnable examples

(package with numerous Shiny apps)

Hello all:

I am planning to submit a package to Bioconductor and have one last
item from BiocCheck (error, warning, note) I have been unable to
resolve:

ERROR: At least 80% of man pages documenting exported objects must
have runnable examples. The following pages do not: plotLitreApp.Rd,
plotPCPApp.Rd, plotSMApp.Rd, plotVolcanoApp.Rd

I have 18 man pages (9 function-related, 8 data-related, and 1
package-related). Of these, 4 of the function-related man pages (the
ones listed in the ERROR) are Shiny applications of the following
format:

appDir <- system.file("shiny-examples", "plotLitreApp", package =

"bigPint")

shiny::runApp(appDir, display.mode = "normal")

If I do not have \dontrun{} around these shiny app examples, then R
CMD check permanently halts on the "checking examples..." step. If I
do have \dontrun{} around these shiny app examples, then R CMD
BiocCheck gives me the error above. My question is: What is the
recommended procedure in such a situation where the package is being
prepared for Bioconductor submission?

An example of one script causing the error can be found at:
https://github.com/lrutter/bigPint/blob/master/R/plotLitreApp.R

Thank you for any advice!
Lindsay

information.  If you are not the intended recipient(s), or the employee or
agent responsible for the delivery of this message to the intended
recipient(s), you are hereby notified that any disclosure, copying,
distribution, or use of this email message is prohibited.  If you have
received this message in error, please notify the sender immediately by
e-mail and delete this email message from your computer. Thank you.

        [[alternative HTML version deleted]]