Julia 1.0 Programming Cookbook

How to do it...

In order to parallelize computations over DataFrame, follow these steps:

1. In the first step, we import the Julia packages required for the computation:

using DataFrames, BenchmarkTools, Random, Statistics

2. This recipe will be illustrated by an artificial dataset with 100,000 rows filled with random values:

Random.seed!(0);N = 100_000;const data = DataFrame(rowtype = rand(1:12, N));data.x1 = data.rowtype .* randn(N);

3. Let's examine the beginning of the data (since we use Random.seed!(0), you should see an identical output):

julia> head(data, 5)5×2 DataFrame│ Row   │ rowtype │ x1       ││       │ Int64   │ Float64  ││-------│---------│----------││ 1     │ 1       │ -0.298115││ 2     │ 3       │ -2.71766 ││ 3     │ 10      │ -28.0064 ││ 4     │ 6       │ 4.39991  ││ 5 │ 1 │ ...