Skip to content
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

Optimize completecases to process only missingable columns #2726

Merged
merged 7 commits into from
May 7, 2021
Merged

Optimize completecases to process only missingable columns #2726

Show file tree
Hide file tree
Changes from 5 commits
Commits
Show all changes
7 commits
Select commit Hold shift + click to select a range
e642415
completecases process only missingable cols
pstorozenko Apr 17, 2021
c4e07f7
Split oprations in completecases
pstorozenko Apr 18, 2021
79e34df
Some tests for completecases added
pstorozenko May 2, 2021
7bb1ee5
Comment added
pstorozenko May 2, 2021
0cbf724
Update test/data.jl
bkamins May 2, 2021
1f2fb7f
Apply suggestions from code review
bkamins May 3, 2021
920bf79
Update test/data.jl
bkamins May 5, 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
20 changes: 16 additions & 4 deletions src/abstractdataframe/abstractdataframe.jl
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -761,14 +761,26 @@ function completecases(df::AbstractDataFrame, col::Colon=:)
"data frame with no columns"))
end
res = trues(size(df, 1))
aux = BitVector(undef, size(df, 1))
for i in 1:size(df, 2)
res .&= .!ismissing.(df[!, i])
v = df[!, i]
if Missing <: eltype(v)
# Disable fused broadcasting as it happens to be much slower
aux .= .!ismissing.(v)
res .&= aux
end
end
res
return res
end

completecases(df::AbstractDataFrame, col::ColumnIndex) =
.!ismissing.(df[!, col])
function completecases(df::AbstractDataFrame, col::ColumnIndex)
v = df[!, col]
if Missing <: eltype(v)
return .!ismissing.(v)
else
return trues(size(df, 1))
end
end

completecases(df::AbstractDataFrame, cols::MultiColumnIndex) =
completecases(df[!, cols])
Expand Down
14 changes: 12 additions & 2 deletions test/data.jl
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -114,8 +114,17 @@ end
:auto)
df2 = DataFrame([Union{Int, Missing}[1, 2, 3, 4], ["one", "two", missing, "four"]],
:auto)

@test df2[completecases(df2), :] == df2[[1, 2, 4], :]
df3 = DataFrame(x = Int[1, 2, 3, 4], y = Union{Int, Missing}[1, missing, 2, 3],
z = Missing[missing, missing, missing, missing])

@test completecases(df2), :] == .!ismissing.(df2.x2)
@test @inferred(completecases(df3, :x)) == trues(nrow(df3))
@test completecases(df3, :y) == .!ismissing.(df3.y)
@test completecases(df3, :z) == completecases(df3, [:z, :x]) ==
completecases(df3, [:x, :z]) == completecases(df3, [:y, :x, :z]) ==
falses(nrow(df3))
@test @inferred(completecases(df3, [:y, :x])) ==
completecases(df3, [:x, :y]) == .!ismissing.(df3.y)
@test dropmissing(df2) == df2[[1, 2, 4], :]
returned = dropmissing(df1)
@test df1 == returned && df1 !== returned
Expand All @@ -128,6 +137,7 @@ end

@test_throws ArgumentError completecases(DataFrame())
@test_throws MethodError completecases(DataFrame(x=1), true)
@test_throws ArgumentError completecases(df3, :a)

for cols in (:x2, "x2", [:x2], ["x2"], [:x1, :x2], ["x1", "x2"], 2, [2], 1:2,
[true, true], [false, true], :,
Expand Down