Adding progress bar to functions that do not imply iterations

[angela_italy]

Hi,
I would like to know whether there's way to add a progress bar to a general function with does not imply iterating over multiple elements .

I had a look at the package progressr which provides tools to add a progress bar but I couldn't find any examples on functions that do not require iterations.

E.g. adding a progress bar to the join_progress funcion hereunder

Thank you for your help

library(sf)
#> Linking to GEOS 3.10.2, GDAL 3.3.2, PROJ 7.2.1; sf_use_s2() is TRUE
#> WARNING: different compile-time and runtime versions for GEOS found:
#> Linked against: 3.10.2-CAPI-1.16.0 compiled against: 3.9.1-CAPI-1.14.2
#> It is probably a good idea to reinstall sf, and maybe rgeos and rgdal too
library(raster)
#> Loading required package: sp
library(dplyr)
#> 
#> Attaching package: 'dplyr'
#> The following objects are masked from 'package:raster':
#> 
#>     intersect, select, union
#> The following objects are masked from 'package:stats':
#> 
#>     filter, lag
#> The following objects are masked from 'package:base':
#> 
#>     intersect, setdiff, setequal, union
library(spData)
#> To access larger datasets in this package, install the spDataLarge
#> package with: `install.packages('spDataLarge',
#> repos='https://nowosad.github.io/drat/', type='source')`
library(progressr)

#  enable the global progress handler

handlers(global = TRUE)
#> Error in globalCallingHandlers(condition = global_progression_handler): should not be called with handlers on the stack



set.seed(2018) # set seed for reproducibility
(bb_world = st_bbox(world)) # the world's bounds
#>       xmin       ymin       xmax       ymax 
#> -180.00000  -89.90000  179.99999   83.64513
#>   xmin   ymin   xmax   ymax 
#> -180.0  -89.9  180.0   83.6
random_df = tibble(
  x = runif(n = 10, min = bb_world[1], max = bb_world[3]),
  y = runif(n = 10, min = bb_world[2], max = bb_world[4])
)
random_points = random_df %>% 
  st_as_sf(coords = c("x", "y")) %>% # set coordinates
  st_set_crs(4326) # set geographic CRS

world_random = world[random_points, ]



join_progress<-function(x) {
  
  Sys.sleep(10)
  st_join(x, world["name_long"])
  
}


random_joined = join_progress(random_points)

[AlexisW]

Short answer: not possible to do what you want.

Longer answer, it's easy to do what you ask: the iteration is not necessary, you could add calls to a progressor between other function calls:

library(progressr)


my_function_without_iteration <- function(){
  p <- progressr::progressor(steps = 5)
  
  # First long operation
  Sys.sleep(.2)
  p()
  # Second long operation
  Sys.sleep(.2)
  p()
  # Third long operation
  Sys.sleep(.2)
  p()
  # Fourth long operation
  Sys.sleep(.2)
  p()
  # Fifth long operation
  Sys.sleep(.2)
  
}

with_progress(my_function_without_iteration())

^{Created on 2023-12-15 with reprex v2.0.2}

However to do that you still need to insert the calls to p() inside the computation. In your example code, there are two long steps Sys.sleep(10) and st_join(x, world["name_long"]), and you have no idea how long they will take to process (well, you do in the case of Sys.sleep but you wouldn't for a real computation). How would a progressor know when to display progress?

This only works if you have access to the inside of the function. So unless the developers who wrote st_join() included code to manage the progress bar, you can't just slap it on top.

[angela_italy]

@AlexisW

with_progress(my_function_without_iteration())

Thank you,
so I should call p after every step like this

join_progress<-function(x) {
  p<-progressr::progressor(step=2)
  Sys.sleep(10)
  p()
  st_join(x, world["name_long"])
  p()
}


random_joined = with_progress(join_progress(random_points))

[AlexisW]

Yes, that would give you a bar that progresses twice, 50% after the Sys.sleep(), and 100% after the st_join().