Extending slackr

Progress Bars

Create text progress bar that is sent directly to a Slack channel.

progress_slackr <- function(x){
  text_slackr('0%')
  s <- c(0,x,1)
  for(i in 1:length(s)){
    Sys.sleep(0.5)
    edit_slackr(sprintf('%s%%',round(100*i/length(s),1)))
  }
  Sys.sleep(0.5)
}

set.seed(1234)
progress_slackr(sort(runif(10)))

This also solves a long standing problem of tracking the progress of parallel jobs on a server. We use slackr with the qapply package, which runs jobs on an Open Grid Scheduler/Engine. We can track each node

Onexit

Attach on.exit expressions to any function in R and at the end of the original function an output will be sent to the Slack channel.

This is useful for letting you know when a simulation is done, but also to be able to send to the Slack channel a relevant summary to see that the simulation did as intended.

ctl <- c(4.17,5.58,5.18,6.11,4.50,4.61,5.17,4.53,5.33,5.14)
trt <- c(4.81,4.17,4.41,3.59,5.87,3.83,6.03,4.89,4.32,4.69)
group <- gl(2, 10, 20, labels = c("Ctl","Trt"))
weight <- c(ctl, trt)

#pass a message to Slack channel 'general'
register_onexit(lm,'bazinga!',channel="#general")

lm.D9 <- slack_lm(weight ~ group)

#test that output keeps inheritance
summary(lm.D9)

#pass a message to Slack channel 'general' with a header message to begin output
register_onexit(lm,'bazinga!',
channel="#general",
header_msg='This is a message to begin')

lm.D9 <- slack_lm(weight ~ group)

#onexit with an expression that calls lm.plot
register_onexit(lm,{
 par(mfrow = c(2, 2), oma = c(0, 0, 2, 0))
 plot(z)
},
channel="#general",
header_msg = 'This is a plot just for this output',
use_device = TRUE)

lm.D9 <- slack_lm(weight ~ group)

#clean up slack channel from examples
delete_slackr(count = 6,channel = '#general')