Skip to content

Improve quantile performance v2 #91

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
wants to merge 6 commits into
base: master
Choose a base branch
from
Open

Improve quantile performance v2 #91

Changes from 3 commits
Commits
Show all changes
6 commits
Select commit Hold shift + click to select a range
b6ceb83
Improve quantile performance v2
bkamins Oct 17, 2021
7943e97
Update src/Statistics.jl
bkamins Oct 17, 2021
af561f2
Update src/Statistics.jl
bkamins Oct 17, 2021
6e233de
Update src/Statistics.jl
bkamins Oct 17, 2021
d84deda
Update runtests.jl
bkamins Oct 17, 2021
fb4c8d8
Update runtests.jl
bkamins Oct 17, 2021
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
26 changes: 12 additions & 14 deletions src/Statistics.jl
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -937,8 +937,7 @@ function quantile!(q::AbstractArray, v::AbstractVector, p::AbstractArray;
end
isempty(q) && return q

minp, maxp = extrema(p)
_quantilesort!(v, sorted, minp, maxp)
_quantilesort!(v, sorted, p)

for (i, j) in zip(eachindex(p), eachindex(q))
@inbounds q[j] = _quantile(v,p[i], alpha=alpha, beta=beta)
Expand All @@ -949,35 +948,34 @@ end
function quantile!(v::AbstractVector, p::Union{AbstractArray, Tuple{Vararg{Real}}};
sorted::Bool=false, alpha::Real=1., beta::Real=alpha)
if !isempty(p)
minp, maxp = extrema(p)
_quantilesort!(v, sorted, minp, maxp)
_quantilesort!(v, sorted, p isa Tuple ? collect(p) : p)
end
return map(x->_quantile(v, x, alpha=alpha, beta=beta), p)
end

quantile!(v::AbstractVector, p::Real; sorted::Bool=false, alpha::Real=1., beta::Real=alpha) =
_quantile(_quantilesort!(v, sorted, p, p), p, alpha=alpha, beta=beta)
_quantile(_quantilesort!(v, sorted, [p]), p, alpha=alpha, beta=beta)

# Function to perform partial sort of v for quantiles in given range
function _quantilesort!(v::AbstractArray, sorted::Bool, minp::Real, maxp::Real)
function _quantilesort!(v::AbstractArray, sorted::Bool, p::AbstractVector{<:Real})
isempty(v) && throw(ArgumentError("empty data vector"))
require_one_based_indexing(v)

if !sorted
lv = length(v)
lo = floor(Int,minp*(lv))
hi = ceil(Int,1+maxp*(lv))

# only need to perform partial sort
sort!(v, 1, lv, Base.Sort.PartialQuickSort(lo:hi), Base.Sort.Forward)
start = 1
for pv in sort(p)
lv = length(v)
Copy link
Member

@nalimilan nalimilan Oct 17, 2021

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Move this out of the loop? BTW, better use lastindex even if we call require_one_based_indexing(v).

lo = floor(Int,pv*(lv))
hi = ceil(Int,1+pv*(lv))
sort!(v, start, lv, Base.Sort.PartialQuickSort(lo:hi), Base.Sort.Forward)
start = hi + 1
Copy link
Member

@nalimilan nalimilan Oct 17, 2021

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Are you completely sure of the +1? Is that still correct if p contains duplicates? That would be worth testing.

end
end
if (sorted && (ismissing(v[end]) || (v[end] isa Number && isnan(v[end])))) ||
any(x -> ismissing(x) || (x isa Number && isnan(x)), v)
throw(ArgumentError("quantiles are undefined in presence of NaNs or missing values"))
end
return v
end

# Core quantile lookup function: assumes `v` sorted
@inline function _quantile(v::AbstractVector, p::Real; alpha::Real=1.0, beta::Real=alpha)
0 <= p <= 1 || throw(ArgumentError("input probability out of [0,1] range"))
Expand Down