Progress bars using broadcast notation in Julia?-CodePudding

In Julia, some amazing packages like ProgressMeter provide an easy API for displaying progress bars in the REPL. However, it seems like they only work with regular for loops, comprehensions and map/reduce. So, e.g., these work:

using ProgressMeter

function sleepyAdd(arg::int64)
    sleep(1)
    return arg   1000
end

nums = [1, 2, 3, 4]

# Progress bar with for loop
@showprogress for num in nums
    sleepyAdd(num)
end

# Progress bar with comprehension
@showprogress [sleepyAdd(num) for num in nums]

Howeve, you can't just use that with broadcast notation (so @showprogress sleepyAdd.(nums) doesn't work), which is a shame because brodcasting functions is one of the best features of Julia's syntax: elegant, simple and concise.

Is there any way around this limitation?

CodePudding user response：

The convenient macro @showprogress doesn't work in this case. But doing @macroexpand @showprogress... gives what is needed to get a working version for specific uses (and maybe for someone to update ProgressMeter.jl):

let t_nums = nums, 
    t_meter = ProgressMeter.Progress(ProgressMeter.length(t_nums)),
    t_wrapper = ProgressMeter.ProgressWrapper(t_nums, t_meter)
    # now comes the comprehension
    [sleepyAdd(num) for num in t_wrapper]
end

And you get the same effect as the for loop.

CodePudding user response：

I don't think there is - here's a link to the only mention of broadcasting in ProgressMeter's issues:

https://github.com/timholy/ProgressMeter.jl/issues/31#issuecomment-448558683

in which Tim says (in 2018) that he's not aware of anyone working on it.