iamed2

# if you know what Blue Prince is but don't know what numerical cores are, don't worry about it

# not particularly optimized, but certainly fast enough

# please comment with a test case if there are bugs

using Combinatorics

const OPERATOR_OPTIONS = map(Tuple, permutations((-,*,/)))

iamed2

function tally(n=100)
    snake_eyes = 0
    any_ones = 0

    for i in 1:n
        rolls = (dice_roll(), dice_roll())
        if 1 in rolls  # tests at least one of them is a one
            any_ones += 1

            if rolls[1] == rolls[2]  # tests both are one

iamed2

julia --startup-file=no --history-file=no -e 'import REPL; print(
let SYMBOLS = REPL.REPLCompletions.latex_symbols
	get(SYMBOLS, "\\\$(ARGS[1])") do
		SYMBOLS["\\text$(ARGS[1])"]
	end
end
)' Sigma

iamed2

# Originally written to work with InlineString in WeakRefStrings, this should work for any primitive type

# In most cases, the built-in defaults provided by Julia work fine, using `read` and `write`.
# However, if your `read` and `write` methods are not round-trippable, or they convert your data into a 
# less-efficient representation before writing, then you might want to additionally define a custom 
# serialization method, using this code.
# It should only be necessary to substitute your type for `Primitive`.

using Serialization

iamed2

using Pkg

function imports(pkg_dir::AbstractString)
    imported = Set{String}()
    for (root, dirs, files) in walkdir(pkg_dir)
        for file in files
            if endswith(file, ".jl")
                union!(imported, open(imports, joinpath(root, file)))
            end
        end

iamed2

julia> @benchmark collect(groupby2(first, ["face", "foo", "bar", "book", "baz"]))
BenchmarkTools.Trial:
  memory estimate:  1.98 KiB
  allocs estimate:  60
  --------------
  minimum time:     14.258 μs (0.00% GC)
  median time:      15.473 μs (0.00% GC)
  mean time:        16.915 μs (0.00% GC)
  maximum time:     207.366 μs (0.00% GC)
  --------------

iamed2

Benchmarking Take1 ...
  Take1 with cfg = 4096: nruns = 28955, elapsed = 1.516598519 secs
  Take1 with cfg = 16384: nruns = 6217, elapsed = 1.307575089 secs
  Take1 with cfg = 65536: nruns = 1252, elapsed = 1.152236431 secs
  Take1 with cfg = 262144: nruns = 266, elapsed = 0.905467819 secs
  Take1 with cfg = 1048576: nruns = 91, elapsed = 1.172819189 secs
Benchmarking Take2 ...
  Take2 with cfg = 4096: nruns = 24779, elapsed = 1.368126555 secs
  Take2 with cfg = 16384: nruns = 6403, elapsed = 1.47713414 secs
  Take2 with cfg = 65536: nruns = 926, elapsed = 0.909415618 secs

iamed2

module Sukano
using JuMP

function sudoku_setup()
    m = Model()

    @defVar(m, 0 <= board[1:9,1:9,1:9] <=1, Int)

    # each square can only have one value
    @addConstraint(m, squareconst[i=1:9,j=1:9], sum(board[i,j,:]) == 1)

iamed2

function partition(f::Function, collection)
    u = similar(collection)
    v = similar(collection)

    for el in collection
        if f(el)
            push!(u, el)
        else
            push!(v, el)
        end

iamed2

data {
	int<lower=1> N_ages;
	vector[N_ages] ages;
	int K[N_ages];
	int N[N_ages];
}

transformed data {
	vector[N_ages] mu;
	mu <- rep_vector(0, N_ages);