Possible pitfalls

using DataFrames
using BenchmarkTools

Know what is copied when creating a DataFrame

x = DataFrame(rand(3, 5), :auto)

3×5 DataFrame

Row	x1	x2	x3	x4	x5
	Float64	Float64	Float64	Float64	Float64
1	0.102373	0.715619	0.402129	0.663427	0.79705
2	0.62435	0.402815	0.0745343	0.972842	0.811803
3	0.421527	0.980137	0.0678962	0.937666	0.789266

x and y are not the same object

y = copy(x)
x === y

false

x and y are not the same object

y = DataFrame(x)
x === y

false

the columns are also not the same

any(x[!, i] === y[!, i] for i in ncol(x))

false

x and y are not the same object

y = DataFrame(x, copycols=false)
x === y

false

But the columns are the same

all(x[!, i] === y[!, i] for i in ncol(x))

true

the same when creating data frames using kwarg syntax

x = 1:3;
y = [1, 2, 3];
df = DataFrame(x=x, y=y)

3×2 DataFrame

Row	x	y
	Int64	Int64
1	1	1
2	2	2
3	3	3

different object

y === df.y

false

range is converted to a vector

typeof(x), typeof(df.x)

(UnitRange{Int64}, Vector{Int64})

slicing rows always creates a copy

y === df[:, :y]

false

you can avoid copying by using copycols=false keyword argument in functions.

df = DataFrame(x=x, y=y, copycols=false)

3×2 DataFrame

Row	x	y
	Int64	Int64
1	1	1
2	2	2
3	3	3

now it is the same

y === df.y

true

not the same object

select(df, :y)[!, 1] === y

false

the same object

select(df, :y, copycols=false)[!, 1] === y

true

Do not modify the parent of GroupedDataFrame or view

x = DataFrame(id=repeat([1, 2], outer=3), x=1:6)
g = groupby(x, :id)

x[1:3, 1] = [2, 2, 2]
g ## well - it is wrong now, g is only a view

GroupedDataFrame with 2 groups based on key: id

First Group (3 rows): id = 2

Row	id	x
	Int64	Int64
1	2	1
2	2	3
3	1	5

⋮

Last Group (3 rows): id = 2

Row	id	x
	Int64	Int64
1	2	2
2	2	4
3	2	6

s = view(x, 5:6, :)

2×2 SubDataFrame

Row	id	x
	Int64	Int64
1	1	5
2	2	6

delete!(x, 3:6)

2×2 DataFrame

Row	id	x
	Int64	Int64
1	2	1
2	2	2

This is an error

s ## Will return BoundsError

Single column selection for a DataFrame

Single column selection for a DataFrame creates aliases with ! and getproperty syntax and copies with :

x = DataFrame(a=1:3)
x.b = x[!, 1] ## alias
x.c = x[:, 1] ## copy
x.d = x[!, 1][:] ## copy
x.e = copy(x[!, 1]) ## explicit copy
display(x)

3×5 DataFrame

Row	a	b	c	d	e
	Int64	Int64	Int64	Int64	Int64
1	1	1	1	1	1
2	2	2	2	2	2
3	3	3	3	3	3

x[1, 1] = 100
display(x)

3×5 DataFrame

Row	a	b	c	d	e
	Int64	Int64	Int64	Int64	Int64
1	100	100	1	1	1
2	2	2	2	2	2
3	3	3	3	3	3

When iterating rows of a data frame

• use eachrow to avoid compilation cost in wide tables,

• but Tables.namedtupleiterator for fast execution in tall tables

The table below is tall:

df2 = DataFrame(rand(10^6, 10), :auto)

1000000×10 DataFrame

999975 rows omitted

Row	x1	x2	x3	x4	x5	x6	x7	x8	x9	x10
	Float64	Float64	Float64	Float64	Float64	Float64	Float64	Float64	Float64	Float64
1	0.766072	0.958625	0.832073	0.0254966	0.847788	0.314068	0.162517	0.264803	0.100111	0.0251814
2	0.369232	0.698069	0.816405	0.581132	0.907493	0.128747	0.603082	0.550548	0.943366	0.657636
3	0.37886	0.688995	0.318009	0.954047	0.534389	0.803777	0.481564	0.0241329	0.341135	0.660792
4	0.365775	0.176099	0.790104	0.685668	0.426212	0.164362	0.903828	0.271555	0.64879	0.901741
5	0.747757	0.465838	0.522257	0.794617	0.166702	0.771483	0.801306	0.138768	0.685883	0.230728
6	0.590686	0.234883	0.403878	0.611071	0.388746	0.281314	0.944573	0.729575	0.295263	0.875675
7	0.0693997	0.441801	0.747364	0.75684	0.829098	0.172671	0.797226	0.0123324	0.0525178	0.981163
8	0.682463	0.719511	0.873292	0.138598	0.22788	0.704053	0.596968	0.123704	0.929795	0.213875
9	0.759339	0.626111	0.622908	0.520235	0.637596	0.658228	0.0669593	0.917496	0.729016	0.827624
10	0.830865	0.658525	0.136859	0.327927	0.72403	0.181743	0.617011	0.658946	0.945421	0.543078
11	0.45224	0.950252	0.524913	0.636485	0.896756	0.526336	0.375906	0.155847	0.109402	0.700583
12	0.706518	0.548551	0.442289	0.637226	0.925868	0.679232	0.670576	0.434754	0.321455	0.0329374
13	0.397241	0.836678	0.212289	0.994819	0.379538	0.0108845	0.198273	0.65096	0.0799598	0.37601
⋮	⋮	⋮	⋮	⋮	⋮	⋮	⋮	⋮	⋮	⋮
999989	0.404831	0.405713	0.229537	0.525418	0.451297	0.943304	0.481654	0.0567747	0.178495	0.0693898
999990	0.349499	0.388355	0.881326	0.0624467	0.487766	0.69541	0.74954	0.623448	0.201285	0.807722
999991	0.418072	0.331013	0.651961	0.229347	0.18391	0.399454	0.0548266	0.216161	0.973492	0.0173772
999992	0.238997	0.0435125	0.0930842	0.268393	0.802257	0.257628	0.918723	0.23682	0.332129	0.936689
999993	0.438404	0.385233	0.336612	0.677567	0.171174	0.253147	0.093383	0.46	0.74405	0.948417
999994	0.787164	0.713589	0.858848	0.615152	0.712799	0.890109	0.915295	0.510148	0.615832	0.626102
999995	0.354889	0.692322	0.739987	0.489757	0.588558	0.163898	0.118598	0.70944	0.905823	0.0660051
999996	0.858621	0.068783	0.807774	0.301436	0.176622	0.943327	0.35847	0.0714865	0.684629	0.593146
999997	0.168081	0.772569	0.0557187	0.272444	0.868914	0.689224	0.440164	0.185854	0.849483	0.0965424
999998	0.805764	0.0757029	0.452915	0.579394	0.433996	0.992273	0.436371	0.128252	0.676505	0.743283
999999	0.235839	0.799897	0.583111	0.401753	0.703255	0.931781	0.642978	0.297659	0.475144	0.0156555
1000000	0.471203	0.00555274	0.718645	0.540562	0.519271	0.264605	0.659582	0.499696	0.315777	0.956476

@time map(sum, eachrow(df2));

  2.248705 seconds (60.04 M allocations: 1.053 GiB, 10.63% gc time, 2.38% compilation time)

@time map(sum, eachrow(df2));

  2.415540 seconds (59.99 M allocations: 1.050 GiB, 3.73% gc time)

@time map(sum, Tables.namedtupleiterator(df2));

  0.295434 seconds (371.24 k allocations: 27.132 MiB, 4.34% gc time, 91.02% compilation time)

@time map(sum, Tables.namedtupleiterator(df2));

  0.014394 seconds (25 allocations: 7.631 MiB)

as you can see - this time it is much faster to iterate a type stable container still you might want to use the select syntax, which is optimized for such reductions:

this includes compilation time

@time select(df2, AsTable(:) => ByRow(sum) => "sum").sum

  0.013925 seconds (6.67 k allocations: 7.969 MiB, 51.19% compilation time)

1000000-element Vector{Float64}:
 4.2967365440983585
 6.255710306108731
 5.185700498816111
 5.334134450728015
 5.325339163319068
 5.355664387891267
 4.860413103161113
 5.210139354465969
 6.365514164469623
 5.62440555470397
 ⋮
 4.128232610040115
 4.50798787807542
 7.245036793462261
 4.829276873210631
 4.864294862813554
 4.398993642137623
 5.324455470285535
 5.0870718591125135
 4.951369427152281

Do it again

@time select(df2, AsTable(:) => ByRow(sum) => "sum").sum

  0.007440 seconds (127 allocations: 7.634 MiB)

1000000-element Vector{Float64}:
 4.2967365440983585
 6.255710306108731
 5.185700498816111
 5.334134450728015
 5.325339163319068
 5.355664387891267
 4.860413103161113
 5.210139354465969
 6.365514164469623
 5.62440555470397
 ⋮
 4.128232610040115
 4.50798787807542
 7.245036793462261
 4.829276873210631
 4.864294862813554
 4.398993642137623
 5.324455470285535
 5.0870718591125135
 4.951369427152281

---

This notebook was generated using Literate.jl.