Mean with weight view fails

[adkabo]

using StatsBase, Statistics
julia> mean(rand(100), @view ProbabilityWeights(rand(100))[:])
ERROR: MethodError: objects of type Vector{Float64} are not callable
Use square brackets [] for indexing an Array.
Stacktrace:
 [1] _mean(f::Vector{Float64}, A::SubArray{Float64, 1, ProbabilityWeights{Float64, Float64, Vector{Float64}}, Tuple{UnitRange{Int64}}, false}, dims::Colon)
   @ Statistics /nix/store/15kh8pp59zbnhxcxh2l66xr1hzly00y9-julia-bin-1.7.1/share/julia/stdlib/v1.7/Statistics/src/Statistics.jl:176
 [2] mean(f::Vector{Float64}, A::SubArray{Float64, 1, ProbabilityWeights{Float64, Float64, Vector{Float64}}, Tuple{UnitRange{Int64}}, false}; dims::Function)
   @ Statistics /nix/store/15kh8pp59zbnhxcxh2l66xr1hzly00y9-julia-bin-1.7.1/share/julia/stdlib/v1.7/Statistics/src/Statistics.jl:104
 [3] mean(f::Vector{Float64}, A::SubArray{Float64, 1, ProbabilityWeights{Float64, Float64, Vector{Float64}}, Tuple{UnitRange{Int64}}, false})
   @ Statistics /nix/store/15kh8pp59zbnhxcxh2l66xr1hzly00y9-julia-bin-1.7.1/share/julia/stdlib/v1.7/Statistics/src/Statistics.jl:104
 [4] top-level scope
   @ REPL[7]:1

julia> versioninfo()
Julia Version 1.7.1
Commit ac5cc99908 (2021-12-22 19:35 UTC)

[ararslan]

This is happening because StatsBase provides weighting by defining methods like mean(x, weights::AbstractWeights) but using @view makes weights a SubArray, so method that's called is mean(f, input) which is interpreting the vector as a function to call. The issue here is actually with StatsBase, which should probably be defining view/Base.maybeview on AbstractWeights to return something that is also <:AbstractWeights.

[nalimilan]

See #723.