Learn R Programming

spatstat.geom (version 3.2-5)

progressreport: Print Progress Reports

Description

Prints Progress Reports during a loop or iterative calculation.

Usage

progressreport(i, n,
               every = min(100,max(1, ceiling(n/100))),
               tick = 1,
               nperline = NULL,
               charsperline = getOption("width"),
               style = spatstat.options("progress"),
               showtime = NULL,
               state=NULL,
               formula = (time ~ i),
               savehistory=FALSE)

Value

If state was NULL, the result is NULL. Otherwise the result is the updated value of state.

Arguments

i

Integer. The current iteration number (from 1 to n).

n

Integer. The (maximum) number of iterations to be computed.

every

Optional integer. Iteration number will be printed when i is a multiple of every.

tick

Optional integer. A tick mark or dot will be printed when i is a multiple of tick.

nperline

Optional integer. Number of iterations per line of output.

charsperline

Optional integer. The number of characters in a line of output.

style

Character string determining the style of display. Options are "tty" (the default), "tk" and "txtbar". See Details.

showtime

Optional. Logical value indicating whether to print the estimated time remaining. Applies only when style="tty".

state

Optional. A list containing the internal data.

formula

Optional. A model formula expressing the expected relationship between the iteration number i and the clock time time. Used for predicting the time remaining.

savehistory

Optional. Logical value indicating whether to save the elapsed times at which progressreport was called.

Author

Adrian Baddeley Adrian.Baddeley@curtin.edu.au, Rolf Turner r.turner@auckland.ac.nz and Ege Rubak rubak@math.aau.dk.

Details

This is a convenient function for reporting progress during an iterative sequence of calculations or a suite of simulations.

  • If style="tk" then tcltk::tkProgressBar is used to pop-up a new graphics window showing a progress bar. This requires the package tcltk. As i increases from 1 to n, the bar will lengthen. The arguments every, tick, nperline, showtime are ignored.

  • If style="txtbar" then txtProgressBar is used to represent progress as a bar made of text characters in the R interpreter window. As i increases from 1 to n, the bar will lengthen. The arguments every, tick, nperline, showtime are ignored.

  • If style="tty" (the default), then progress reports are printed to the console. This only seems to work well under Linux. As i increases from 1 to n, the output will be a sequence of dots (one dot for every tick iterations), iteration numbers (printed when iteration number is a multiple of every or is less than 4), and optionally the estimated time remaining and the estimated completion time.

    The estimated time remaining will be printed only if style="tty", and the argument state is given, and either showtime=TRUE, or showtime=NULL and the iterations are slow (defined as: the estimated time remaining is longer than 3 minutes, or the average time per iteration is longer than 20 seconds).

    The estimated completion time will be printed only if the estimated time remaining is printed and the remaining time is longer than 10 minutes.

    By default, the estimated time remaining is calculated by assuming that each iteration takes the same amount of time, and extrapolating. Alternatively, if the argument formula is given, then it should be a model formula, stating the expected relationship between the iteration number i and the clock time time. This model will be fitted to the history of clock times recorded so far, and used to predict the time remaining. (The default formula states that clock time is a linear function of the iteration number, which is equivalent to assuming that each iteration takes the same amount of time.)

It is optional, but strongly advisable, to use the argument state to store and update the internal data for the progress reports (such as the cumulative time taken for computation) as shown in the last example below. This avoids conflicts with other programs that might be calling progressreport at the same time.

Examples

Run this code
  for(i in 1:40) {
     #
     # code that does something...
     # 
     progressreport(i, 40)
  }

  # saving internal state: *recommended*
  sta <- list()
  for(i in 1:20) {
     # some code ...
     sta <- progressreport(i, 20, state=sta)
  }

  #' use text progress bar
  sta <- list()
  for(i in 1:10) {
     # some code ...
     sta <- progressreport(i, 10, state=sta, style="txtbar")
  }

Run the code above in your browser using DataLab