[AUTO] Update precompiles

[github-actions]

Automated changes by create-pull-request GitHub action

[daschw]

On julia 1.4.2:
Current release:

julia> @time using Plots
 10.849055 seconds (23.34 M allocations: 1.206 GiB, 3.90% gc time)

julia> @time display(plot(rand(10)))
 13.588981 seconds (28.49 M allocations: 1.406 GiB, 3.64% gc time)

Now:

julia> @time using Plots
 10.036745 seconds (23.96 M allocations: 1.240 GiB, 3.97% gc time)

julia> @time display(plot(rand(10)))
 11.619033 seconds (12.34 M allocations: 611.958 MiB, 1.76% gc time)

[BeastyBlacksmith]

what about 1.5?

[daschw]

will check later today

[timholy]

For me,

julia 1.4 (downloaded binary):

julia> @time using Plots; @time p = plot(rand(10)); @time display(p)
 10.592077 seconds (25.83 M allocations: 1.327 GiB, 4.69% gc time)
  3.981842 seconds (4.98 M allocations: 243.521 MiB, 1.94% gc time)
  6.818551 seconds (7.36 M allocations: 367.920 MiB, 1.43% gc time)

julia 1.5rc1 (built from source):

julia> @time using Plots; @time p = plot(rand(10)); @time display(p)
  7.181116 seconds (15.59 M allocations: 883.141 MiB, 3.48% gc time)
  2.843696 seconds (4.57 M allocations: 230.298 MiB, 2.56% gc time)
  4.355263 seconds (6.78 M allocations: 347.737 MiB, 2.23% gc time)

julia master with JuliaLang/julia#35877 (and the customization in JuliaLang/julia#35877 (comment)), JuliaLang/julia#36449, JuliaLang/julia#36452, and JuliaLang/julia#36459:

julia> @time using Plots; @time p = plot(rand(10)); @time display(p)
  5.503256 seconds (10.61 M allocations: 654.482 MiB, 4.07% gc time)
  2.660786 seconds (3.74 M allocations: 204.619 MiB, 2.43% gc time)
  4.071296 seconds (6.06 M allocations: 336.316 MiB, 2.41% gc time)

All three are Plots 1.5.1 and with Revise running in the background. (Without Revise, 1.4 is almost 2s faster in total; on 1.5 and nightly, Revise doesn't cause a performance hit.)

[timholy]

Inspired by #2782, for me here's 1.2 (which is Plots 1.4.4):

julia> @time using Plots; @time p = plot(rand(10)); @time display(p)
  7.298756 seconds (15.10 M allocations: 867.473 MiB, 6.12% gc time)
  2.944357 seconds (5.40 M allocations: 275.429 MiB, 3.80% gc time)
  8.641176 seconds (21.83 M allocations: 1.090 GiB, 5.30% gc time)

with Revise, and

julia> @time using Plots; @time p = plot(rand(10)); @time display(p)
  6.444962 seconds (12.18 M allocations: 719.997 MiB, 6.11% gc time)
  2.946809 seconds (5.08 M allocations: 259.969 MiB, 1.84% gc time)
  6.196895 seconds (8.80 M allocations: 449.288 MiB, 1.58% gc time)

without.

So nightly is by far the fastest yet.