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
  3.827 ns (0 allocations: 0 bytes)
@btime $x.x500;  ## Slower
  15.630 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`
  149.432 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();
  3.681 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();
  3.165 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();
  3.763 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();
  29.541 ms (1949728 allocations: 37.40 MiB)
  28.865 ms (1950028 allocations: 45.03 MiB)
  1.142 ms (728 allocations: 7.66 MiB)
  1.527 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
  2.477 ms (212 allocations: 38.16 MiB)
  1.124 μs (29 allocations: 1.50 KiB)
  432.462 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:
  1.812 ms (8 allocations: 7.63 MiB)
 categorical:
  15.993 ms (1000004 allocations: 30.52 MiB)
String
 raw:
  21.960 ms (4 allocations: 448 bytes)
 categorical:
  32.089 ms (1000004 allocations: 30.52 MiB)
Union{Missing, Int64}
 raw:
  6.031 ms (4 allocations: 464 bytes)
 categorical:
  15.935 ms (1000004 allocations: 30.52 MiB)
Union{Missing, String}
 raw:
  22.191 ms (4 allocations: 448 bytes)
 categorical:
  33.176 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 (2500724 rows): x = 'b': ASCII/Unicode U+0062 (category Ll: Letter, lowercase)
2500699 rows omitted
Rowxy
CharInt64
1b1
2b1
3b1
4b1
5b1
6b1
7b1
8b1
9b1
10b1
11b1
12b1
13b1
2500713b1
2500714b1
2500715b1
2500716b1
2500717b1
2500718b1
2500719b1
2500720b1
2500721b1
2500722b1
2500723b1
2500724b1

Last Group (2500304 rows): x = 'c': ASCII/Unicode U+0063 (category Ll: Letter, lowercase)
2500279 rows omitted
Rowxy
CharInt64
1c1
2c1
3c1
4c1
5c1
6c1
7c1
8c1
9c1
10c1
11c1
12c1
13c1
2500293c1
2500294c1
2500295c1
2500296c1
2500297c1
2500298c1
2500299c1
2500300c1
2500301c1
2500302c1
2500303c1
2500304c1

traditional syntax, slow

@btime combine(v -> sum(v.y), $gdf)
  15.846 ms (332 allocations: 19.09 MiB)
4×2 DataFrame
Rowxx1
CharInt64
1b2500724
2d2500237
3a2498735
4c2500304

use column selector

@btime combine($gdf, :y => sum)
  6.916 ms (198 allocations: 10.14 KiB)
4×2 DataFrame
Rowxy_sum
CharInt64
1b2500724
2d2500237
3a2498735
4c2500304
transform!(df, :x => categorical => :x);
gdf = groupby(df, :x)

GroupedDataFrame with 4 groups based on key: x

First Group (2498735 rows): x = CategoricalArrays.CategoricalValue{Char, UInt32} 'a'
2498710 rows omitted
Rowxy
Cat…Int64
1a1
2a1
3a1
4a1
5a1
6a1
7a1
8a1
9a1
10a1
11a1
12a1
13a1
2498724a1
2498725a1
2498726a1
2498727a1
2498728a1
2498729a1
2498730a1
2498731a1
2498732a1
2498733a1
2498734a1
2498735a1

Last Group (2500237 rows): x = CategoricalArrays.CategoricalValue{Char, UInt32} 'd'
2500212 rows omitted
Rowxy
Cat…Int64
1d1
2d1
3d1
4d1
5d1
6d1
7d1
8d1
9d1
10d1
11d1
12d1
13d1
2500226d1
2500227d1
2500228d1
2500229d1
2500230d1
2500231d1
2500232d1
2500233d1
2500234d1
2500235d1
2500236d1
2500237d1
@btime combine($gdf, :y => sum)
  6.866 ms (206 allocations: 10.62 KiB)
4×2 DataFrame
Rowxy_sum
Cat…Int64
1a2498735
2b2500724
3c2500304
4d2500237
transform!(df, :x => PooledArray{Char} => :x)
10000000×2 DataFrame
9999975 rows omitted
Rowxy
CharInt64
1b1
2b1
3d1
4a1
5c1
6b1
7c1
8b1
9a1
10c1
11a1
12a1
13d1
9999989d1
9999990a1
9999991c1
9999992d1
9999993c1
9999994d1
9999995b1
9999996b1
9999997b1
9999998a1
9999999a1
10000000a1
gdf = groupby(df, :x)

GroupedDataFrame with 4 groups based on key: x

First Group (2500724 rows): x = 'b': ASCII/Unicode U+0062 (category Ll: Letter, lowercase)
2500699 rows omitted
Rowxy
CharInt64
1b1
2b1
3b1
4b1
5b1
6b1
7b1
8b1
9b1
10b1
11b1
12b1
13b1
2500713b1
2500714b1
2500715b1
2500716b1
2500717b1
2500718b1
2500719b1
2500720b1
2500721b1
2500722b1
2500723b1
2500724b1

Last Group (2500304 rows): x = 'c': ASCII/Unicode U+0063 (category Ll: Letter, lowercase)
2500279 rows omitted
Rowxy
CharInt64
1c1
2c1
3c1
4c1
5c1
6c1
7c1
8c1
9c1
10c1
11c1
12c1
13c1
2500293c1
2500294c1
2500295c1
2500296c1
2500297c1
2500298c1
2500299c1
2500300c1
2500301c1
2500302c1
2500303c1
2500304c1
@btime combine($gdf, :y => sum)
  6.882 ms (200 allocations: 10.20 KiB)
4×2 DataFrame
Rowxy_sum
CharInt64
1b2500724
2d2500237
3a2498735
4c2500304

Use views instead of materializing a new DataFrame#

x = DataFrame(rand(100, 1000), :auto)
@btime $x[1:1, :]
  177.661 μs (3993 allocations: 159.07 KiB)
1×1000 DataFrame
900 columns omitted
Rowx1x2x3x4x5x6x7x8x9x10x11x12x13x14x15x16x17x18x19x20x21x22x23x24x25x26x27x28x29x30x31x32x33x34x35x36x37x38x39x40x41x42x43x44x45x46x47x48x49x50x51x52x53x54x55x56x57x58x59x60x61x62x63x64x65x66x67x68x69x70x71x72x73x74x75x76x77x78x79x80x81x82x83x84x85x86x87x88x89x90x91x92x93x94x95x96x97x98x99x100
Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64
10.9500580.5874940.1779940.6552750.9733260.4099070.8359260.6491910.2027920.8088650.5812850.340270.1367580.6607410.975270.5793350.909860.8488910.5756170.6274890.2902920.7286370.02112940.8209520.8465810.1323860.1079260.4304220.9611990.7295270.592140.27480.3346640.5242270.823970.01763950.465530.1353550.7800290.587380.4838120.655280.5848150.07787160.5373810.4362960.03058090.4625070.5053820.1705840.2219430.5359460.784980.7309940.415150.4180.4505290.4897280.7917030.4653040.264860.0679450.5084960.1847480.4615170.1935330.3870850.9318580.4250180.09519190.2063340.9184690.7752170.8658250.2488520.4300640.2918650.1000740.7396420.6850140.09495250.697680.4665320.7411470.5636580.8653010.5401110.3598680.03935460.5405050.7709540.8739070.09777180.860350.5679510.8919520.09675760.3454340.7317820.21978
@btime $x[1, :]
  22.833 ns (0 allocations: 0 bytes)
DataFrameRow (1000 columns)
900 columns omitted
Rowx1x2x3x4x5x6x7x8x9x10x11x12x13x14x15x16x17x18x19x20x21x22x23x24x25x26x27x28x29x30x31x32x33x34x35x36x37x38x39x40x41x42x43x44x45x46x47x48x49x50x51x52x53x54x55x56x57x58x59x60x61x62x63x64x65x66x67x68x69x70x71x72x73x74x75x76x77x78x79x80x81x82x83x84x85x86x87x88x89x90x91x92x93x94x95x96x97x98x99x100
Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64
10.9500580.5874940.1779940.6552750.9733260.4099070.8359260.6491910.2027920.8088650.5812850.340270.1367580.6607410.975270.5793350.909860.8488910.5756170.6274890.2902920.7286370.02112940.8209520.8465810.1323860.1079260.4304220.9611990.7295270.592140.27480.3346640.5242270.823970.01763950.465530.1353550.7800290.587380.4838120.655280.5848150.07787160.5373810.4362960.03058090.4625070.5053820.1705840.2219430.5359460.784980.7309940.415150.4180.4505290.4897280.7917030.4653040.264860.0679450.5084960.1847480.4615170.1935330.3870850.9318580.4250180.09519190.2063340.9184690.7752170.8658250.2488520.4300640.2918650.1000740.7396420.6850140.09495250.697680.4665320.7411470.5636580.8653010.5401110.3598680.03935460.5405050.7709540.8739070.09777180.860350.5679510.8919520.09675760.3454340.7317820.21978
@btime view($x, 1:1, :)
  21.908 ns (0 allocations: 0 bytes)
1×1000 SubDataFrame
900 columns omitted
Rowx1x2x3x4x5x6x7x8x9x10x11x12x13x14x15x16x17x18x19x20x21x22x23x24x25x26x27x28x29x30x31x32x33x34x35x36x37x38x39x40x41x42x43x44x45x46x47x48x49x50x51x52x53x54x55x56x57x58x59x60x61x62x63x64x65x66x67x68x69x70x71x72x73x74x75x76x77x78x79x80x81x82x83x84x85x86x87x88x89x90x91x92x93x94x95x96x97x98x99x100
Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64
10.9500580.5874940.1779940.6552750.9733260.4099070.8359260.6491910.2027920.8088650.5812850.340270.1367580.6607410.975270.5793350.909860.8488910.5756170.6274890.2902920.7286370.02112940.8209520.8465810.1323860.1079260.4304220.9611990.7295270.592140.27480.3346640.5242270.823970.01763950.465530.1353550.7800290.587380.4838120.655280.5848150.07787160.5373810.4362960.03058090.4625070.5053820.1705840.2219430.5359460.784980.7309940.415150.4180.4505290.4897280.7917030.4653040.264860.0679450.5084960.1847480.4615170.1935330.3870850.9318580.4250180.09519190.2063340.9184690.7752170.8658250.2488520.4300640.2918650.1000740.7396420.6850140.09495250.697680.4665320.7411470.5636580.8653010.5401110.3598680.03935460.5405050.7709540.8739070.09777180.860350.5679510.8919520.09675760.3454340.7317820.21978
@btime $x[1:1, 1:20]
  3.850 μs (70 allocations: 3.09 KiB)
1×20 DataFrame
Rowx1x2x3x4x5x6x7x8x9x10x11x12x13x14x15x16x17x18x19x20
Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64
10.9500580.5874940.1779940.6552750.9733260.4099070.8359260.6491910.2027920.8088650.5812850.340270.1367580.6607410.975270.5793350.909860.8488910.5756170.627489
@btime $x[1, 1:20]
  22.531 ns (0 allocations: 0 bytes)
DataFrameRow (20 columns)
Rowx1x2x3x4x5x6x7x8x9x10x11x12x13x14x15x16x17x18x19x20
Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64
10.9500580.5874940.1779940.6552750.9733260.4099070.8359260.6491910.2027920.8088650.5812850.340270.1367580.6607410.975270.5793350.909860.8488910.5756170.627489
@btime view($x, 1:1, 1:20)
  23.893 ns (0 allocations: 0 bytes)
1×20 SubDataFrame
Rowx1x2x3x4x5x6x7x8x9x10x11x12x13x14x15x16x17x18x19x20
Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64Float64
10.9500580.5874940.1779940.6552750.9733260.4099070.8359260.6491910.2027920.8088650.5812850.340270.1367580.6607410.975270.5793350.909860.8488910.5756170.627489

This notebook was generated using Literate.jl.