Performance tips#

using DataFrames
using BenchmarkTools
using CategoricalArrays
using PooledArrays
using Random

Access by column number is faster than by name#

x = DataFrame(rand(5, 1000), :auto)
@btime $x[!, 500]; ## Faster
  3.085 ns (0 allocations: 0 bytes)
@btime $x.x500;  ## Slower
  11.121 ns (0 allocations: 0 bytes)

When working with data DataFrame use barrier functions or type annotation#

function f_bad() ## this function will be slow
    Random.seed!(1)
    x = DataFrame(rand(1000000, 2), :auto)
    y, z = x[!, 1], x[!, 2]
    p = 0.0
    for i in 1:nrow(x)
        p += y[i] * z[i]
    end
    p
end

@btime f_bad();
# if you run @code_warntype f_bad() then you notice
# that Julia does not know column types of `DataFrame`
  105.709 ms (5999022 allocations: 122.06 MiB)

solution 1 is to use barrier function (it should be possible to use it in almost any code) for the calculation. You will notice much less memopry allocations and faster performance.

function f_inner(y, z)
    p = 0.0
    for i in eachindex(y, z)
        p += y[i] * z[i]
    end
    p
end

function f_barrier()
    Random.seed!(1)
    x = DataFrame(rand(1000000, 2), :auto)
    f_inner(x[!, 1], x[!, 2])
end

@btime f_barrier();
  3.819 ms (44 allocations: 30.52 MiB)

or use inbuilt function if possible

using LinearAlgebra

function f_inbuilt()
    Random.seed!(1)
    x = DataFrame(rand(1000000, 2), :auto)
    dot(x[!, 1], x[!, 2])
end

@btime f_inbuilt();
  3.205 ms (44 allocations: 30.52 MiB)

solution 2 is to provide the types of extracted columns. However, there are cases in which you will not know these types.

function f_typed()
    Random.seed!(1)
    x = DataFrame(rand(1000000, 2), :auto)
    y::Vector{Float64}, z::Vector{Float64} = x[!, 1], x[!, 2]
    p = 0.0
    for i in 1:nrow(x)
        p += y[i] * z[i]
    end
    p
end

@btime f_typed();
  3.737 ms (44 allocations: 30.52 MiB)

In general for tall and narrow tables it is often useful to use Tables.rowtable, Tables.columntable or Tables.namedtupleiterator for intermediate processing of data in a type-stable way.

Consider using delayed DataFrame creation technique#

also notice the difference in performance between copying vs non-copying data frame creation

function f1()
    x = DataFrame([Vector{Float64}(undef, 10^4) for i in 1:100], :auto, copycols=false) ## we work with a DataFrame directly
    for c in 1:ncol(x)
        d = x[!, c]
        for r in 1:nrow(x)
            d[r] = rand()
        end
    end
    x
end

function f1a()
    x = DataFrame([Vector{Float64}(undef, 10^4) for i in 1:100], :auto) ## we work with a DataFrame directly
    for c in 1:ncol(x)
        d = x[!, c]
        for r in 1:nrow(x)
            d[r] = rand()
        end
    end
    x
end

function f2()
    x = Vector{Any}(undef, 100)
    for c in 1:length(x)
        d = Vector{Float64}(undef, 10^4)
        for r in eachindex(d)
            d[r] = rand()
        end
        x[c] = d
    end
    DataFrame(x, :auto, copycols=false) ## we delay creation of DataFrame after we have our job done
end

function f2a()
    x = Vector{Any}(undef, 100)
    for c in eachindex(x)
        d = Vector{Float64}(undef, 10^4)
        for r in eachindex(d)
            d[r] = rand()
        end
        x[c] = d
    end
    DataFrame(x, :auto) ## we delay creation of DataFrame after we have our job done
end

@btime f1();
@btime f1a();
@btime f2();
@btime f2a();
  29.051 ms (1949728 allocations: 37.40 MiB)
  29.817 ms (1950028 allocations: 45.03 MiB)
  1.145 ms (728 allocations: 7.66 MiB)
  1.596 ms (1028 allocations: 15.29 MiB)

You can add rows to a DataFrame in place and it is fast#

x = DataFrame(rand(10^6, 5), :auto)
y = DataFrame(transpose(1.0:5.0), :auto)
z = [1.0:5.0;]

@btime vcat($x, $y); ## creates a new DataFrame - slow
@btime append!($x, $y); ## in place - fast

x = DataFrame(rand(10^6, 5), :auto) ## reset to the same starting point
@btime push!($x, $z); ## add a single row in place - fast
  2.527 ms (212 allocations: 38.16 MiB)
  1.174 μs (29 allocations: 1.50 KiB)
  440.874 ns (16 allocations: 256 bytes)

Allowing missing as well as categorical slows down computations#

using StatsBase

function test(data) ## uses countmap function to test performance
    println(eltype(data))
    x = rand(data, 10^6)
    y = categorical(x)
    println(" raw:")
    @btime countmap($x)
    println(" categorical:")
    @btime countmap($y)
    nothing
end

test(1:10)
test([randstring() for i in 1:10])
test(allowmissing(1:10))
test(allowmissing([randstring() for i in 1:10]))
Int64
 raw:
  1.890 ms (8 allocations: 7.63 MiB)
 categorical:
  15.323 ms (1000004 allocations: 30.52 MiB)
String
 raw:
  21.787 ms (4 allocations: 448 bytes)
 categorical:
  32.270 ms (1000004 allocations: 30.52 MiB)
Union{Missing, Int64}
 raw:
  6.025 ms (4 allocations: 464 bytes)
 categorical:
  15.817 ms (1000004 allocations: 30.52 MiB)
Union{Missing, String}
 raw:
  21.463 ms (4 allocations: 448 bytes)
 categorical:
  32.639 ms (1000004 allocations: 30.52 MiB)

When aggregating use column selector and prefer integer, categorical, or pooled array grouping variable#

df = DataFrame(x=rand('a':'d', 10^7), y=1);
gdf = groupby(df, :x)

GroupedDataFrame with 4 groups based on key: x

First Group (2499891 rows): x = 'd': ASCII/Unicode U+0064 (category Ll: Letter, lowercase)
2499866 rows omitted
Rowxy
CharInt64
1d1
2d1
3d1
4d1
5d1
6d1
7d1
8d1
9d1
10d1
11d1
12d1
13d1
2499880d1
2499881d1
2499882d1
2499883d1
2499884d1
2499885d1
2499886d1
2499887d1
2499888d1
2499889d1
2499890d1
2499891d1

Last Group (2501420 rows): x = 'b': ASCII/Unicode U+0062 (category Ll: Letter, lowercase)
2501395 rows omitted
Rowxy
CharInt64
1b1
2b1
3b1
4b1
5b1
6b1
7b1
8b1
9b1
10b1
11b1
12b1
13b1
2501409b1
2501410b1
2501411b1
2501412b1
2501413b1
2501414b1
2501415b1
2501416b1
2501417b1
2501418b1
2501419b1
2501420b1

traditional syntax, slow

@btime combine(v -> sum(v.y), $gdf)
  16.437 ms (332 allocations: 19.09 MiB)
4×2 DataFrame
Rowxx1
CharInt64
1d2499891
2c2499309
3a2499380
4b2501420

use column selector

@btime combine($gdf, :y => sum)
  6.901 ms (198 allocations: 10.14 KiB)
4×2 DataFrame
Rowxy_sum
CharInt64
1d2499891
2c2499309
3a2499380
4b2501420
transform!(df, :x => categorical => :x);
gdf = groupby(df, :x)

GroupedDataFrame with 4 groups based on key: x

First Group (2499380 rows): x = CategoricalArrays.CategoricalValue{Char, UInt32} 'a'
2499355 rows omitted
Rowxy
Cat…Int64
1a1
2a1
3a1
4a1
5a1
6a1
7a1
8a1
9a1
10a1
11a1
12a1
13a1
2499369a1
2499370a1
2499371a1
2499372a1
2499373a1
2499374a1
2499375a1
2499376a1
2499377a1
2499378a1
2499379a1
2499380a1

Last Group (2499891 rows): x = CategoricalArrays.CategoricalValue{Char, UInt32} 'd'
2499866 rows omitted
Rowxy
Cat…Int64
1d1
2d1
3d1
4d1
5d1
6d1
7d1
8d1
9d1
10d1
11d1
12d1
13d1
2499880d1
2499881d1
2499882d1
2499883d1
2499884d1
2499885d1
2499886d1
2499887d1
2499888d1
2499889d1
2499890d1
2499891d1
@btime combine($gdf, :y => sum)
  6.843 ms (206 allocations: 10.62 KiB)
4×2 DataFrame
Rowxy_sum
Cat…Int64
1a2499380
2b2501420
3c2499309
4d2499891
transform!(df, :x => PooledArray{Char} => :x)
10000000×2 DataFrame
9999975 rows omitted
Rowxy
CharInt64
1d1
2c1
3c1
4a1
5c1
6a1
7c1
8c1
9d1
10c1
11d1
12d1
13d1
9999989c1
9999990c1
9999991b1
9999992a1
9999993d1
9999994a1
9999995d1
9999996c1
9999997c1
9999998d1
9999999a1
10000000d1
gdf = groupby(df, :x)

GroupedDataFrame with 4 groups based on key: x

First Group (2499891 rows): x = 'd': ASCII/Unicode U+0064 (category Ll: Letter, lowercase)
2499866 rows omitted
Rowxy
CharInt64
1d1
2d1
3d1
4d1
5d1
6d1
7d1
8d1
9d1
10d1
11d1
12d1
13d1
2499880d1
2499881d1
2499882d1
2499883d1
2499884d1
2499885d1
2499886d1
2499887d1
2499888d1
2499889d1
2499890d1
2499891d1

Last Group (2501420 rows): x = 'b': ASCII/Unicode U+0062 (category Ll: Letter, lowercase)
2501395 rows omitted
Rowxy
CharInt64
1b1
2b1
3b1
4b1
5b1
6b1
7b1
8b1
9b1
10b1
11b1
12b1
13b1
2501409b1
2501410b1
2501411b1
2501412b1
2501413b1
2501414b1
2501415b1
2501416b1
2501417b1
2501418b1
2501419b1
2501420b1
@btime combine($gdf, :y => sum)
  6.905 ms (200 allocations: 10.20 KiB)
4×2 DataFrame
Rowxy_sum
CharInt64
1d2499891
2c2499309
3a2499380
4b2501420

Use views instead of materializing a new DataFrame#

x = DataFrame(rand(100, 1000), :auto)
@btime $x[1:1, :]
  178.674 μs (3993 allocations: 159.07 KiB)
1×1000 DataFrame
900 columns omitted
Rowx1x2x3x4x5x6x7x8x9x10x11x12x13x14x15x16x17x18x19x20x21x22x23x24x25x26x27x28x29x30x31x32x33x34x35x36x37x38x39x40x41x42x43x44x45x46x47x48x49x50x51x52x53x54x55x56x57x58x59x60x61x62x63x64x65x66x67x68x69x70x71x72x73x74x75x76x77x78x79x80x81x82x83x84x85x86x87x88x89x90x91x92x93x94x95x96x97x98x99x100
Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64
10.5691790.3392930.8051340.8168530.8719540.2848040.1822570.9181690.2687910.4523610.5375720.1778280.9387770.2681840.186880.7919020.01074450.7477370.4329730.261990.05108240.9386310.364470.3312880.0631290.1423110.7477280.344420.5080990.7816170.7083170.3234660.2269840.6792540.7073070.9126940.2421550.4877420.5492960.8817620.7351890.5035390.5871350.454860.9856530.2896270.793860.4087350.1622350.2213840.6583060.4904370.482310.5211420.8325290.7176410.8123720.05131090.2780660.4719860.1473520.1338970.003108680.3759540.2676210.1439850.4919890.2367040.7311150.73150.6467030.8896790.7753760.8882510.1476390.8451120.3650460.4931020.3625950.07148520.001290750.4887940.9762730.7358280.5016550.4094070.7512530.4879380.5922830.7606210.3425840.3113550.1457550.6129370.04760680.3754460.3178370.3663980.9444310.605746
@btime $x[1, :]
  22.230 ns (0 allocations: 0 bytes)
DataFrameRow (1000 columns)
900 columns omitted
Rowx1x2x3x4x5x6x7x8x9x10x11x12x13x14x15x16x17x18x19x20x21x22x23x24x25x26x27x28x29x30x31x32x33x34x35x36x37x38x39x40x41x42x43x44x45x46x47x48x49x50x51x52x53x54x55x56x57x58x59x60x61x62x63x64x65x66x67x68x69x70x71x72x73x74x75x76x77x78x79x80x81x82x83x84x85x86x87x88x89x90x91x92x93x94x95x96x97x98x99x100
Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64
10.5691790.3392930.8051340.8168530.8719540.2848040.1822570.9181690.2687910.4523610.5375720.1778280.9387770.2681840.186880.7919020.01074450.7477370.4329730.261990.05108240.9386310.364470.3312880.0631290.1423110.7477280.344420.5080990.7816170.7083170.3234660.2269840.6792540.7073070.9126940.2421550.4877420.5492960.8817620.7351890.5035390.5871350.454860.9856530.2896270.793860.4087350.1622350.2213840.6583060.4904370.482310.5211420.8325290.7176410.8123720.05131090.2780660.4719860.1473520.1338970.003108680.3759540.2676210.1439850.4919890.2367040.7311150.73150.6467030.8896790.7753760.8882510.1476390.8451120.3650460.4931020.3625950.07148520.001290750.4887940.9762730.7358280.5016550.4094070.7512530.4879380.5922830.7606210.3425840.3113550.1457550.6129370.04760680.3754460.3178370.3663980.9444310.605746
@btime view($x, 1:1, :)
  22.230 ns (0 allocations: 0 bytes)
1×1000 SubDataFrame
900 columns omitted
Rowx1x2x3x4x5x6x7x8x9x10x11x12x13x14x15x16x17x18x19x20x21x22x23x24x25x26x27x28x29x30x31x32x33x34x35x36x37x38x39x40x41x42x43x44x45x46x47x48x49x50x51x52x53x54x55x56x57x58x59x60x61x62x63x64x65x66x67x68x69x70x71x72x73x74x75x76x77x78x79x80x81x82x83x84x85x86x87x88x89x90x91x92x93x94x95x96x97x98x99x100
Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64
10.5691790.3392930.8051340.8168530.8719540.2848040.1822570.9181690.2687910.4523610.5375720.1778280.9387770.2681840.186880.7919020.01074450.7477370.4329730.261990.05108240.9386310.364470.3312880.0631290.1423110.7477280.344420.5080990.7816170.7083170.3234660.2269840.6792540.7073070.9126940.2421550.4877420.5492960.8817620.7351890.5035390.5871350.454860.9856530.2896270.793860.4087350.1622350.2213840.6583060.4904370.482310.5211420.8325290.7176410.8123720.05131090.2780660.4719860.1473520.1338970.003108680.3759540.2676210.1439850.4919890.2367040.7311150.73150.6467030.8896790.7753760.8882510.1476390.8451120.3650460.4931020.3625950.07148520.001290750.4887940.9762730.7358280.5016550.4094070.7512530.4879380.5922830.7606210.3425840.3113550.1457550.6129370.04760680.3754460.3178370.3663980.9444310.605746
@btime $x[1:1, 1:20]
  3.810 μs (70 allocations: 3.09 KiB)
1×20 DataFrame
Rowx1x2x3x4x5x6x7x8x9x10x11x12x13x14x15x16x17x18x19x20
Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64
10.5691790.3392930.8051340.8168530.8719540.2848040.1822570.9181690.2687910.4523610.5375720.1778280.9387770.2681840.186880.7919020.01074450.7477370.4329730.26199
@btime $x[1, 1:20]
  21.909 ns (0 allocations: 0 bytes)
DataFrameRow (20 columns)
Rowx1x2x3x4x5x6x7x8x9x10x11x12x13x14x15x16x17x18x19x20
Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64
10.5691790.3392930.8051340.8168530.8719540.2848040.1822570.9181690.2687910.4523610.5375720.1778280.9387770.2681840.186880.7919020.01074450.7477370.4329730.26199
@btime view($x, 1:1, 1:20)
  22.532 ns (0 allocations: 0 bytes)
1×20 SubDataFrame
Rowx1x2x3x4x5x6x7x8x9x10x11x12x13x14x15x16x17x18x19x20
Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64
10.5691790.3392930.8051340.8168530.8719540.2848040.1822570.9181690.2687910.4523610.5375720.1778280.9387770.2681840.186880.7919020.01074450.7477370.4329730.26199

This notebook was generated using Literate.jl.