Uniform zeros like syntax

[aminya]

julia> Uniform(Float32,-1,1)
Uniform{Float32}(a=-1.0f0, b=1.0f0)

[codecov-io]

Codecov Report

Merging #1035 into master will decrease coverage by 0.01%.
The diff coverage is 0%.

@@            Coverage Diff             @@
##           master    #1035      +/-   ##
==========================================
- Coverage   78.27%   78.26%   -0.02%     
==========================================
  Files         112      112              
  Lines        5345     5346       +1     
==========================================
  Hits         4184     4184              
- Misses       1161     1162       +1

Impacted Files	Coverage Δ
src/univariate/continuous/uniform.jl	84% <0%> (-1.72%)	⬇️

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 3151573...5130723. Read the comment docs.

[matbesancon]

The feature is interesting. Wasn't there anything already in the package doing this?
Before merging that, could you add a bunch of tests and docstring to Uniform explaining the behaviour?

[andreasnoack]

I think Uniform{Float32}(-1, 1) would be more in line with how we typically do this, e.g.

julia> Complex{Float32}(1,1)
1.0f0 + 1.0f0im

[aminya]

I think Uniform{Float32}(-1, 1) would be more in line with how we typically do this, e.g.

julia> Complex{Float32}(1,1)
1.0f0 + 1.0f0im

That is one of the APIs that I specified in this issue However, we need to first fix the type stability. The issue is bigger than the API

[aminya]

Proceeded by #1045