particle_simulation.jl

### A Pluto.jl notebook ###
# v0.16.0

using Markdown
using InteractiveUtils

# ╔═╡ a82f22e4-f35b-461a-b481-1dff43722e44
using StaticArrays

# ╔═╡ d42f842d-6c2a-40db-b0c4-e936244a9e7c
using BenchmarkTools

# ╔═╡ 99b5c818-a825-4939-849e-1cade802f63d
using Measurements

# ╔═╡ d6564250-f646-40de-9463-a956af1a5b1d
using ForwardDiff

# ╔═╡ d2f4d622-8e35-4be3-b421-39b28a748cab
using CellListMap

# ╔═╡ f049ab19-7ecf-4c65-bf6d-21352c1fe767
using FastPow

# ╔═╡ 7c792b6b-b6ee-4e30-88d5-d0b8064f2734
begin
    using Plots
    plot_font = "Computer Modern"
    default(
        fontfamily=plot_font,
        linewidth=2, framestyle=:box, label=:none, grid=false,
        size=(400,400)
    )    
end

# ╔═╡ febe8c06-b3aa-4db1-a3ea-fdc2a81bdebd
using Printf

# ╔═╡ a756dd18-fac6-4527-944e-c16d8cc4bf95
begin
    using PlutoUI
    TableOfContents()
end

# ╔═╡ a87fad48-73c1-4a08-a6f1-aae759b3c6fc
md"""
# Particle Simulations with Julia

Leandro Martínez

Institute of Chemistry - University of Campinas

[http://m3g.iqm.unicamp.br](http://m3g.iqm.unicamp.br) - 
[https://github.com/m3g](https://github.com/m3g)

"""

# ╔═╡ 172227c2-b27a-40db-91f4-9566c2f6cf52
md"""
# Outline

- Elements of a particle simulation
- Benchmarking vs. a conventional compiled language (Fortran)
- Exploring the generic character of functions
- Differentiable simulations and parameter fitting
- Using cell lists
- An efficient and generic cell list implementation
- The Packmol strategy
- Benchmarking vs. NAMD
- Remarks

"""

# ╔═╡ 4b484cf6-4888-4f04-b3fd-94862822b0c0
md"""
# Defining the type of particle

We define a simple vector in 2D space, with coordinates `x` and `y`. The vector will be defined with the aid of the `StaticArrays` package, which provides convenient constructors for this type of variable, and all the arithmetics. The memory layout of a vector of these vectors is identical to that of a `N×M` matrix, where `N` is the dimensio nf the space (2D here) and `M` is the number of vector. Julia is column-major, thus this is the most efficient memory layout for this type of computation.
"""

# ╔═╡ 8c444ee4-8c77-413a-bbeb-9e5ae2428876
struct Vec2D{T} <: FieldVector{2,T}
    x::T
    y::T
end

# ╔═╡ a0de01b5-779a-48c0-8d61-12b02a5f527e
md"""
For convenience, here we will also define a function that returns a random vector, given a range of coordinates:
"""

# ╔═╡ 532eb3bc-5522-4348-afa5-5336ec6752c7
md"""
In defining the function above we took care of making it generic for the type and dimension of the vectors desired, such that we do not need to redefine it later when peforming simulations with different data structures. 
"""

# ╔═╡ dc5f7484-3dc3-47a7-ad4a-30f97fc14d11
md"""
## Force between a pair of particles 

Initially, the energy function will be a soft potential, which is zero for distances greater than a cutoff, and increasing quadratically for distances smaller than the cutoff:

If $d = ||\vec{y}-\vec{x}||$ is the norm of the relative position of two positions, we have:

$$u(\vec{x},\vec{y},c)=
\begin{cases}
(d-c)^2 &\textrm{if} & d\leq c \\
0 & \textrm{if} & d > c \\
\end{cases}$$

for which the forces are

$$\vec{f_x}(\vec{x},\vec{y},c)=
\begin{cases}
2(d-c)\frac{(\vec{y}-\vec{x})}{d} &\textrm{if} & d\leq c \\
\vec{0} & \textrm{if} & d > c \\
\end{cases}$$
and
$$\vec{f_y} = -\vec{f_x}$$.


"""

# ╔═╡ ab3ff21b-bf82-4d8c-abd1-c27418956ed8
md"""
The standard Julia `LinearAlgebra` library provides a `norm` function, and there is no reason not to use it (although a manual definition of the same function can also be easily implemented):
"""

# ╔═╡ 7a1db355-bba9-4322-9fb4-a6d7b7bdd60d
import LinearAlgebra: norm

# ╔═╡ cc49ef04-08d8-42bb-9170-9db64e275a51
md"""
The energy and force functions are clear to read:
"""

# ╔═╡ d00e56f2-9d3a-4dd3-80eb-3201eff39b96
md"""
And for a unidimensional case, with a defined cutoff, look like:
"""

# ╔═╡ b5c09cd3-6063-4a36-96cd-2d128aa11b82
const cutoff = 5.

# ╔═╡ 7719b317-e85b-4583-b401-a8614d4b2373
md"""
The function that will compute the force over all pairs will just *naively* run over all (non-repeated) the pairs. The function `forces!` will receive as a parameter the function that computes the force between pairs, such that this pairwise function can be changed later. 

Inside `forces!`, the `force_pair` function will receive four parameters: the indexes of the particles and their positions. We will use the indexes later. 
"""

# ╔═╡ 144119ad-ab88-4165-883a-f2fc2464a838
md"""
Let us create some particle positions to explain how the function will be called. 
"""

# ╔═╡ dd9e4332-2908-40ef-b461-6b571df56cf4
md"""
The function `force_pair`, will be passed to the function that computes the forces to all pairs as *closure*, which will capture the value of the cutoff. The closure also allows us to ignore the indexes of the particles, which are expected by the inner implementation of the function inside `forces`. For example:
"""

# ╔═╡ 017cd6a8-712d-4ec5-b31f-7f1f507764f2
md"""
The third argument of `forces!` function is a *closure*, which can be read as: it is the function that *given* `(i,j,x,y)`, returns `fₓ(x,y,cutoff)`. Thus, it is consistent with the internal call of `fₓ` of `forces!`, and *closes over* the additional parameter `cutoff` required for the computation. 
"""

# ╔═╡ b5206dd5-1f46-4437-929b-efd68393b12b
md"""
# Performing a particle simulation

Now, given the function that computes the forces, we can perform a particle simulation. We will use a simple Euler integration scheme, and the algorithm will be:

1. Compute forces at time $t$ from positions $x$:
$f(t) = f(x)$

2. Update the positions (using $a = f/m$):
$x(t + dt) = x(t) + v(t)dt + a(t)dt^2/2$

3. Update the velocities:
$v(t+dt) = v(t) + a(t)dt$

4. Goto 1.

## The actual simulation code is as short:
"""

# ╔═╡ eb5dc224-1491-11ec-1cae-d51c93cd292c
function md(
    x0::Vector{T},
    v0::Vector{T},
    mass,dt,nsteps,isave,forces!
) where T
    x = copy(x0)
    v = copy(v0)
    a = similar(x0)
    f = similar(x0)
    trajectory = [ copy(x0) ] # will store the trajectory
    for step in 1:nsteps
        # Compute forces
        forces!(f,x)
        # Accelerations
        @. a = f / mass
        # Update positions
        @. x = x + v*dt + a*dt^2/2
        # Update velocities
        @. v = v + a*dt
        # Save if required
        if mod(step,isave) == 0
            println("Saved trajectory at step: ",step)
            push!(trajectory,copy(x))
        end
    end
    return trajectory
end

# ╔═╡ 594ba1d6-2dae-4f20-9546-f52fac17c2f0
md"""
By using a parametric type of input (i. e. `Vector{T}`) we can guarantee that an error will be thrown if the positions and velocities are not provided as the same type of variable. 

The `@.` notation is very common in Julia and means that the computation will be performed element-wise.
"""

# ╔═╡ 66c7d960-7e05-4613-84e8-2a40fe40dc3d
md"""
## Let us run the simulation!
"""

# ╔═╡ e717a8d9-ccfb-4f89-b2a2-f244f108b48d
md"""
Here we generate random positions and velocities, and use masses equal to `1.0` for all particles.
"""

# ╔═╡ eab7b195-64d5-4587-8687-48a673ab091b
md"""
## Using periodic boundary conditions

Our particles just explode, since they have initial random velocities and there are only repulsive interactions. 

We have a more interesting dynamics if we use periodic boundary conditions. To do so, we will update how the forces are computed.
"""

# ╔═╡ 34dc72dc-4864-47c0-b730-183f67e7aea3
md"""
## Wrapping of coordinates

The following function defines how to wrap the coordinates on the boundaries, for a square or cubic box of side `side`:
"""

# ╔═╡ 02d9bf3b-708c-4293-b198-9043b334ff7e
md"""
This allows writting the force computation now as:
"""

# ╔═╡ 0a5282ee-c88a-4bcc-aca2-477f28e9e04d
md"""
Our box has a side of 100:
"""

# ╔═╡ b2a4a505-47ff-40bb-9a6d-a08d91c53217
const side = 100.

# ╔═╡ fcff6973-012a-40fc-a618-f6262266287a
md"""
To run the simulation with the new periodic forces, we use the same `md` function, just passing the new `fₓ` function in the *closure* definition:
"""

# ╔═╡ 14867ffd-cde5-43f8-8399-01169ee29b73
md"""
A relevant detail here is that we could use the same `fₓ` name for the function, because it receives the `side` of the box as a parameter, and multiple-dispatch then chooses the correct method automaticaly. 
"""

# ╔═╡ c4798182-de75-4b59-8be7-f7cf1051364d
md"""
While plotting the trajectory, we will wrap the coordinates:
"""

# ╔═╡ 22fb0386-d4fa-47b9-ac31-decf2731cbc1
md"""
## Benchmarking
"""

# ╔═╡ 8e23a3ea-3039-4a5f-b37f-c4710153938e
md"""
Benchmarkming in Julia can be done with the `@time` macro or the macros from the `BenchmarkTools` package. Compilation occurs on the first call to each method, and the macros from `BenchmarkTools` discount the compilation time automatically. 
"""

# ╔═╡ 2a3e7257-63ad-4761-beda-cec18b91f99c
md"""

Something of the order of `200ms` and `200KiB` of allocations does not seem bad, but it doesn't mean anything either. What is interesting to point here is just that this code, compared to ahead-of-time compiled language like Fortran, is completely comparable in terms of performance, as [this benchmark](https://github.com/m3g/2021_FortranCon/tree/main/benchmark_vs_fortran) shows. 

"""

# ╔═╡ 49b1f040-929a-4238-acd9-6554757b592c
md"""
# Exploring generics

## Running the simulations in 3D

Not much is needed to just run the simulation in three dimensions. We only need to define our 3D vector:
"""

# ╔═╡ 26d5c6e9-a903-4792-a0e0-dec1a2e86a01
struct Vec3D{T} <: FieldVector{3,T}
    x::T
    y::T
    z::T
end

# ╔═╡ 2aef27bd-dea6-4a93-9d0f-b9249c9dd2cd
md"""
That is enough such that we can run the simulations in 3D:
"""

# ╔═╡ b6dcb9a3-59e3-4eae-9399-fb072c704f1a
md"""
## Automatic error propagation

Performing simulations in different dimensions is not the most interesting, or most useful property of generic programming. We can, for example, propagate the error in the positions of the particles, simply by defining a type of particle that carries both the position and the cumulative error. 

A small example of how that can be done is shown. First, we create a type of variable that carries both the coordinates and the uncertainty on the coordinates:
"""

# ╔═╡ e4657169-1bb2-4d4a-ac9d-adc80499d07d
struct MyMeasurement{T}
    x::T
    Δx::T
end

# ╔═╡ 5d395353-5681-4780-983e-902fdb89eaf2
md"""
and we will overload the printing of this variables to make things prettier:
"""

# ╔═╡ e9376a4b-3d60-42eb-8681-cd2bcec13fe8
Base.show(io::IO,m::MyMeasurement) = println(io," $(m.x) ± $(m.Δx)")

# ╔═╡ c5cdd71f-5ded-482f-9205-c13f01a14d0b
m = MyMeasurement(1.0,0.1)

# ╔═╡ a0993a1f-60a6-45b5-815e-676c49a9f049
md"""
Now we define the arithmetics for this type of variable. For example, the sum of two `MyMeasurement`s sums the uncertainties, but so do the subtraction. The other uncertainties are also propagaged linearly, according to the first derivative of their operations relative to the values:
"""

# ╔═╡ 4e7f8db4-b5cc-4a3e-9fa7-e62d8f2a36ac
begin
    import Base: -, +, *, /, ^, sqrt
    +(m1::MyMeasurement,m2::MyMeasurement) = MyMeasurement(m1.x+m2.x,m1.Δx+m2.Δx)
    -(m1::MyMeasurement,m2::MyMeasurement) = MyMeasurement(m1.x-m2.x,m1.Δx+m2.Δx)
    *(α,m::MyMeasurement) = MyMeasurement(α*m.x,sign(α)*α*m.Δx)
    *(m::MyMeasurement,α) = α*m
    /(m::MyMeasurement,α) = inv(α)*m
    sqrt(m::MyMeasurement{T}) where T = MyMeasurement(sqrt(m.x),inv(2*sqrt(m.x))*m.Δx)
    ^(m::MyMeasurement{T},n) where T = MyMeasurement{T}(m.x^n,n*m.x^(n-1)*m.Δx)
end

# ╔═╡ f87e4036-8f82-41c7-90c1-daa5f677488d
function random_vec(::Type{VecType},range) where VecType 
    dim = length(VecType)
    T = eltype(VecType)
    p = VecType(
        range[begin] + rand(T)*(range[end]-range[begin]) for _ in 1:dim
    )
    return p
end

# ╔═╡ df33b999-4a42-4133-bf59-5a65240790cf
function energy(x::T,y::T,cutoff) where T
    Δv = y - x
    d = norm(Δv)
    if d > cutoff
        energy = zero(T)
    else
        energy = (d - cutoff)^2
    end
    return energy
end

# ╔═╡ 0f52365d-34f4-46ed-923e-3ea31c6db0ca
function fₓ(x::T,y::T,cutoff) where T
    Δv = y - x
    d = norm(Δv)
    if d > cutoff
        fₓ = zero(T)
    else
        fₓ = 2*(d - cutoff)*(Δv/d)
    end
    return fₓ
end

# ╔═╡ f58769a6-a656-42a3-8bc6-c204d4cfd897
function forces!(f::Vector{T},x::Vector{T},fₓ::F) where {T,F}
    fill!(f,zero(T))
    n = length(x)
    for i in 1:n-1
        for j in i+1:n
            fᵢ = fₓ(i,j,x[i],x[j])
            f[i] += fᵢ 
            f[j] -= fᵢ
        end
    end
    return f
end

# ╔═╡ beeb3335-5c49-47de-a1d3-3eef5f9479f1
function wrap(x,side)
    x = rem(x,side)
    if x >= side/2
        x -= side
    elseif x < -side/2
        x += side
    end
    return x
end

# ╔═╡ 0967b90d-ac88-476d-a57a-7c38dfa82204
function fₓ(x::T,y::T,cutoff,side) where T
    Δv = wrap.(y - x, side)
    d = norm(Δv)
    if d > cutoff
        fₓ = zero(T)
    else
        fₓ = 2*(d - cutoff)*(Δv/d)
    end
    return fₓ
end

# ╔═╡ 36da3e92-000c-4d4b-9abf-4cd588b3a354
md"""
With such definitions, we can operate over variables of type `MyMeasurement`, propagating automatically the uncertainty along the operations:
"""

# ╔═╡ 70eb2e0a-a5c8-4975-8f6c-589035bea29c
sqrt((2*(m + 4*m)^2/3))

# ╔═╡ b4646a29-3efd-4bd1-bffc-3575559de937
md"""
We can also define a 2D (or 3D) vectors of values with uncertainties, without changing the previous definitions of these:
"""

# ╔═╡ d32743c0-fc80-406f-83c5-4528e439589a
x = Vec2D(MyMeasurement(1.0,0.1),MyMeasurement(2.0,0.2))

# ╔═╡ 98478246-5940-4828-a8f1-9c9fa990676d
md"""
Now operations on this vector propagate the uncertainties of the componentes as well:
"""

# ╔═╡ 310f247e-3fe8-4621-ae0b-b5ee38d2ee89
2*x .+ sqrt.(x)

# ╔═╡ 8267220a-f06e-4761-b310-00f8ba44e4b1
md"""
Propagating uncertainties in more general scenarios requires the definition of other propagation rules. Also, one might want to consider the correlation between variables, which makes the propagation rules more complicated and expensive.

Fortunately, there are some package that provide the error propagation in more general scenarios, by defining the proper progagation rules. 

Here, we use the `Measurements`  package.
"""

# ╔═╡ ce916139-221a-462e-877f-88212663c05e
md"""
### Using `Measurements`
"""

# ╔═╡ 8e2903be-4975-4e14-84ed-6e712f47fe47
md"""
Using `Measurments`  we do not need to change anything in our previous code, but only redefine the content of our vectors, which will now carry in each coordinate the position and the error in the position, accumulated from an initial uncertainty:
"""

# ╔═╡ 418f31bb-81d5-459b-b402-4fd4e3f4ab27
md"""
We need to redefine your initial random vector generator only:
"""

# ╔═╡ 05402cbd-78c6-4234-8680-c351c8c37778
function random_vec(::Type{Vec2D{Measurement{T}}},range,Δ) where T    
    p = Vec2D(
        range[begin] + rand(T)*(range[end]-range[begin]) ± rand()*Δ,
        range[begin] + rand(T)*(range[end]-range[begin]) ± rand()*Δ
    )
    return p
end

# ╔═╡ 356ac5a4-c94e-42cb-a085-0198b29c7e52
x0 = [ random_vec(Vec2D{Float64},(0,100)) for _ in 1:100] 

# ╔═╡ d23b4a92-055e-4ed7-bd46-8a3c59312993
f = similar(x0)

# ╔═╡ e6e29d1e-9a93-49db-a358-6b66f0bc3433
forces!(
    f,
    x0, 
    (i,j,x,y) -> fₓ(x,y,cutoff) # closure
) 

# ╔═╡ 3755a4f3-1842-4de2-965e-d294c06c54c7
trajectory = md((
    x0 = [random_vec(Vec2D{Float64},(-50,50)) for _ in 1:100 ], 
    v0 = [random_vec(Vec2D{Float64},(-1,1)) for _ in 1:100 ], 
    mass = [ 1.0 for _ in 1:100 ],
    dt = 0.1,
    nsteps = 1000,
    isave = 10,
    forces! = (f,x) -> forces!(f,x, (i,j,p1,p2) -> fₓ(p1,p2,cutoff))
)...)

# ╔═╡ 985b4ffb-7964-4b50-8c2f-e5f45f352500
trajectory_periodic = md((
    x0 = [random_vec(Vec2D{Float64},(-50,50)) for _ in 1:100 ], 
    v0 = [random_vec(Vec2D{Float64},(-1,1)) for _ in 1:100 ], 
    mass = [ 10.0 for _ in 1:100 ],
    dt = 0.1,
    nsteps = 1000,
    isave = 10,
    forces! = (f,x) -> forces!(f,x,(i,j,p1,p2) -> fₓ(p1,p2,cutoff,side))
)...)

# ╔═╡ 1ad401b5-20b2-489b-b2aa-92f729b1d725
@benchmark md($(
    x0 = [random_vec(Vec2D{Float64},-50:50) for _ in 1:100 ], 
    v0 = [random_vec(Vec2D{Float64},-1:1) for _ in 1:100 ], 
    mass = [ 1.0 for _ in 1:100 ],
    dt = 0.1,
    nsteps = 1000,
    isave = 10,
    forces! = (f,x) -> forces!(f,x, (i,j,p1,p2) -> fₓ(p1,p2,cutoff,side))
)...)

# ╔═╡ 0546ee2d-b62d-4c7a-8172-ba87b3c1aea4
trajectory_periodic_3D = md((
    x0 = [random_vec(Vec3D{Float64},-50:50) for _ in 1:100 ], 
    v0 = [random_vec(Vec3D{Float64},-1:1) for _ in 1:100 ], 
    mass = [ 1.0 for _ in 1:100 ],
    dt = 0.1,
    nsteps = 1000,
    isave = 10,
    forces! = (f,x) -> forces!(f,x,(i,j,p1,p2) -> fₓ(p1,p2,cutoff,side))
)...)

# ╔═╡ 4e97f24c-c237-4117-bc57-e4e88c8fb8d2
md"""
Which generates random positions carrying an initial uncertainty we defined:
"""

# ╔═╡ b31da90d-7165-42de-b18d-90584affea03
random_vec(Vec2D{Measurement{Float64}},(-50,50),1e-5)

# ╔═╡ 5f37640b-ffd9-4877-a78c-a699b2671919
md"""
That given, the same simulation codes can be used to run the particles simulations while propagating the uncertinties of the coordinates of each position:
"""

# ╔═╡ 1d6eedfd-d013-4557-9cf2-103f8fb7b72a
md"""
The trajectory, of course, looks the same (except that we ran less steps, because propagating the error is expensive):
"""

# ╔═╡ c003a61d-a434-4d7b-9214-5b52aa044248
md"""
But now we have an estimate of the error of the positions, propagated from the initial uncertainty:
"""

# ╔═╡ 63eb391f-0238-434a-bc3a-2fa8ed41448e
md"""
### Planetary motion

Perhaps this is more interesting to see in a planetary trajectory:
"""

# ╔═╡ 7b9bb0fd-34a5-42e1-bc35-7259447b73d0
function gravitational_force(i,j,x,y,mass)
    G = 0.00049823382528 # MKm³ / (10²⁴kg days²)
    dr = y - x
    r = norm(dr)
    return G*mass[i]*mass[j]*dr/r^3
end

# ╔═╡ 6a4e0e2e-75c5-4cab-987d-3d6b62f9bb06
md"""
Note that now we need the indexes of the particles to be able to pass the information of their masses. 

A set of planetary positions and velocities is something that we have to obtain [experimentaly](https://nssdc.gsfc.nasa.gov/planetary/factsheet/). Here, the distance units $10^6$ km, and time is in days. Thus, velocities are in MKm per day.

The uncertainty of the positions will be taken as the diameter of each planet. In this illustrative example we will not add uncertainties to the velcities. 
"""

# ╔═╡ c91862dd-498a-4712-8e3d-b77e088cd470
planets_x0 = [
    Vec2D(  0.0 ±  1.39    , 0. ±  1.39    ), # "Sun"
    Vec2D( 57.9 ±  4.879e-3, 0. ±  4.879e-3), # "Mercury"
    Vec2D(108.2 ± 12.104e-3, 0. ± 12.104e-3), # "Venus"
    Vec2D(149.6 ± 12.756e-3, 0. ± 12.756e-3), # "Earth"
    Vec2D(227.9 ±  6.792e-3, 0. ±  6.792e-3), # "Mars"
]

# ╔═╡ a08d6e6d-ddc4-40aa-b7c4-93ea03191415
planets_v0 = [
    Vec2D(0. ± 0.,   0.0 ± 0.), # "Sun"
    Vec2D(0. ± 0.,  4.10 ± 0.), # "Mercury"
    Vec2D(0. ± 0.,  3.02 ± 0.), # "Venus"
    Vec2D(0. ± 0.,  2.57 ± 0.), # "Earth"
    Vec2D(0. ± 0.,  2.08 ± 0.)  # "Mars"  
]

# ╔═╡ a356e2cc-1cb1-457a-986c-998cf1efe008
md"""
And the masses are given in units of $10^{24}$ kg:
"""

# ╔═╡ 57141f7c-9261-4dc5-98e4-b136a15f86fc
const masses = [ 1.99e6, 0.330, 4.87, 5.97, 0.642 ]

# ╔═╡ 055e32d7-073c-40db-a267-750636b9f786
md"""
Let us see the planets orbiting the sun:
"""

# ╔═╡ aaa97ce4-a5ff-4332-89a2-843cee2e5b6d
trajectory_planets = md((
    x0 = planets_x0, 
    v0 = planets_v0, 
    mass = masses,
    dt = 1, # days
    nsteps = 2*365, # two Earth years
    isave = 1, # save every day
    forces! = (f,x) -> forces!(
        f,x, (i,j,p1,p2) -> gravitational_force(i,j,p1,p2,masses)
    )
)...)

# ╔═╡ 93697e4d-369b-48e9-8b28-a0ff58604d02
md"""
If you are wandering why the errors oscilate, it is because the trajectories are periodic. Whenever all possible trajectories starting from within the uncertainty interval cross each other, the error of the predicted position is independent on the initial coordinates. Thus, the derivative of the position is zero relative to initial position, and so it the propagated uncertainty when using a linear propagation rule.
"""

# ╔═╡ c4344e64-aa22-4328-a97a-71e44bcd289f
md"""
One thing I don't like, though, is that in two years the Earth did not complete two  revolutions around the Sun. Something is wrong with our data. Can we improve that?
"""

# ╔═╡ 827bda6f-87d4-4d36-8d89-f144f4595240
md"""
## We can differentiate everything!

Perhaps astoningshly (at least for me), our simulation is completely differentiable. That means that we can tune the parameters of the simulation, and the data, using optimization algorithms that require derivatives. 

Here we speculate that what was wrong with our data was that the initial position of the Earth was somewhat out of place. That caused the Earth orbit to be slower than it should.

We will define, then, an objective function which returns the displacement of the Earth relative to its initial position (at day one) after two years. Our goal is that after one year the Earth returns to its initial position.
"""

# ╔═╡ 1ff4077a-4742-4c5e-a8d6-c4699469a683
md"""
First, we define a function that executes a simulation of *two years* of an Earth orbit, starting from a given position for the Earth `x` coordinate as a parameter. We will be careful in making all other coordinates of the same type of `x`, so that the generality of the type of variable being used is kept consistent:
"""

# ╔═╡ 4a75498d-8f4e-406f-8b01-f6a5f153919f
function earth_orbit(x::T=149.6,nsteps=2*3650,isave=20) where T
    x0 = [
        Vec2D( zero(T), zero(T)), # "Sun"
        Vec2D(       x, zero(T))  # "Earth"
    ]
    v0 = [ 
        Vec2D( zero(T), zero(T)), # "Sun"
        Vec2D( zero(T), 2.57*one(T)), # "Earth"
    ]
    masses = [ 1.99e6, 5.97 ]
    trajectory = md((
        x0 = x0, 
        v0 = v0, 
        mass = masses,
        dt = 0.1, # days
        nsteps = nsteps, # one Earth year
        isave = isave, 
        forces! = (f,x) -> forces!(f,x, 
            (i,j,p1,p2) -> gravitational_force(i,j,p1,p2,masses)
        )
    )...)
    return trajectory
end

# ╔═╡ 3ae783ce-d06e-4cc2-b8a3-94512e8f1490
md"""
Now we define our objective function, consisting of the norm of the difference between the initial and final coordinates of the Earth after two year (what we want is that the Earth returns to its initial position):
"""

# ╔═╡ 13e7da81-8581-4f32-9fdb-2599dd36a12c
function error_in_orbit(x::T=149.6) where T
	nsteps = 2*3650 # two years
    traj = earth_orbit(x,nsteps,nsteps) # Save last point only
    return norm(traj[end][2]-[x,0.])
end

# ╔═╡ 4870b1f3-3134-4ddc-a59d-fa806b456a23
md"""
We can see that our current data results in a significant error:
"""

# ╔═╡ fda6171c-9675-4f2e-b226-7ccf100529cd
error_in_orbit()

# ╔═╡ a862f8a3-0131-4644-bc90-246bf3120790
md"""
We want to minimize this error, and it turns out that your simulation is fully differentiable. We will use here the `ForwardDiff` automatic differentiation package:
"""

# ╔═╡ eee3ac4b-4ddb-4699-b6e6-f0ffcc562c07
md"""
Which can be used just as it it to compute the derivative of the error in the orbit relative to the initial `x` position of the Earth:
"""

# ╔═╡ 107aec28-ecb5-4007-95e5-25d0a7f0c465
ForwardDiff.derivative(error_in_orbit,149.6)

# ╔═╡ 1394e4c6-c371-47c0-8ca8-f0830d63d8ec
md"""
To minimize the error in the orbit we will write a simple stepest descent algorithm. Many packages are available for optimization, but here we will keep things simpler also to illustrate that writting the optimizer in Julia is a valid alternative:
"""

# ╔═╡ 535716e6-9c1c-4324-a4cd-b1214df3c01d
function gradient_descent(x,f,g,tol,maxtrial)
    itrial = 0
    step = 1.0
    fx = f(x)
    gx = g(x)
    while (abs(gx) > tol) && (itrial < maxtrial) && (step > 1e-10)
        xtrial = x - gx*step
        ftrial = f(xtrial)
        if ftrial > fx
            step = step / 2
        else
            x = xtrial
            fx = ftrial
            gx = g(x)
            step = step * 2
        end
        itrial += 1
    end 
    return x, gx, itrial
end

# ╔═╡ b8edfb4e-6780-4ce7-94c1-4073ff7fa832
md"""
The derivative of our error can be computed by *closing over* the `error_in_orbit` function:
"""

# ╔═╡ b8320f78-323c-49a9-a9f9-2748d19ecb35
error_derivative(x) = ForwardDiff.derivative(error_in_orbit,x)

# ╔═╡ 92737d73-676c-4c96-a321-831ecaf37690
md"""
And now we can call the `gradient_descent` function directly:
"""

# ╔═╡ 931a9c5f-8f91-4e88-956b-50c0efc9c58b
best_x0 = gradient_descent(149.6,error_in_orbit,error_derivative,1e-4,1000)

# ╔═╡ b5b96082-efde-464f-bcd4-f2e0a84befcd
md"""
The result is reasonable: the error in the orbit has significantly being disminished:
"""

# ╔═╡ 7658a32c-d3da-4ec9-9d96-0d30bb18f08c
error_in_orbit(best_x0[1])

# ╔═╡ e61981d5-5448-45e9-81dc-320ac87ba813
md"""
Let us see our trajectory now with the new initial condition:
"""

# ╔═╡ 31e1bb51-c531-4c4a-8634-5caafb7e9e51
earth_traj_0 = earth_orbit(149.6)

# ╔═╡ b0b81da4-6788-45c4-b618-188a02b5e09c
earth_traj_best = earth_orbit(best_x0[1])

# ╔═╡ 47c205c3-ceae-4e12-9ade-753df1608deb
md"""
The dark blue dot is the corrected trajectory, and the light blue dot is the original one. Therefore, we were able to optimize the *initial point* of the trajectory with a gradient-based method. This concept can be used for adjusting parameters in simulations of many kinds (particle simulations or differential equations in general).
"""

# ╔═╡ 826693ff-9a9b-46b1-aeb3-767a5e6f9441
md"""
# Accelerating with CellListMap.jl
"""

# ╔═╡ d231842d-9b7a-4711-b71b-5d54041ebc1f
md"""
[`CellListMap.jl`](https://m3g.github.io/CellListMap.jl/stable/) is package aiming an efficient implementation of [cell lists](https://en.wikipedia.org/wiki/Cell_lists). Cell lists are practical algorithm to reduce the cost of computing short-ranged distances between particles. The package provides a general interface to compute any distance-dependent property, as potential energies and forces, nearest-neighbour lists, distribution functions, etc. It accepts systems with general (triclinic) periodic boundary conditions, in two and three dimensions. 

The most simple cell list algorithm is relatively simple. Many optimizations can be done, however, on the construction of the lists, on the handling of periodic conditions, minimization of the number of unnecessary distance computations, and the parallelization of the construction of the lists and the mapping of the property to be evaluated. 

"""

# ╔═╡ 53cedd26-3742-4c23-a8b8-8a1f2bdfa135
md"""
## The naive algorithm is too slow O(n²)
"""

# ╔═╡ 889f837d-2e26-4261-b276-5fd91efdda6a
md"""
With ~1k, particles, the number of pairs of particles is already of the order of hundreds of thousands. The naive O(n²) algorithm is already too slow. Typical simulations involve tenths of thousands to millions of particles.
"""

# ╔═╡ 670a01e3-82f8-4c7f-8577-852081d91ed7
md"""
Here, we will simulate 1000 particles to start:
"""

# ╔═╡ fce1e3e0-cdf7-453f-b913-964c10fa85a6
const n_large = 1000

# ╔═╡ 69a92ac6-833c-4605-b3d0-9400e4572886
md"""
Our previous system had 100 particles in a square of side 100. We will keep the density constant:
"""

# ╔═╡ 542a9ef5-d9ee-49bd-9d31-61e28b80b5cb
const box_side = sqrt(n_large / (100/100^2))

# ╔═╡ 8bada25c-b586-42b4-851d-232ccca8a456
md"""
We only need to generate the coordinates and run:
"""

# ╔═╡ 7600c6dc-769e-4c77-8526-281a1bcec079
x0_large = [ Vec2D(box_side*rand(),box_side*rand()) for _ in 1:n_large ] 

# ╔═╡ 29dbc47b-3697-4fdf-8f34-890ab4d0cdae
t_naive = @elapsed trajectory_periodic_large = md((
    x0 = x0_large, 
    v0 = [random_vec(Vec2D{Float64},(-1,1)) for _ in 1:n_large ], 
    mass = [ 10.0 for _ in 1:n_large ],
    dt = 0.1,
    nsteps = 1000,
    isave = 10,
    forces! = (f,x) -> forces!(f,x,(i,j,p1,p2) -> fₓ(p1,p2,cutoff,box_side))
)...)

# ╔═╡ 0ee7fc18-f41f-4179-a75e-1e1d56b2db29
md""" 
Running time of naive algorithm: $t_naive seconds
"""

# ╔═╡ 0d0374ed-5150-40e6-b5a4-9a344b6ca47a
md"""
## Using cell lists
"""

# ╔═╡ f7cf613e-be9d-4f62-a778-cc4375eb99df
md"""
In cell lists, the particles are classified in cells before any distance computation. The distances are computed only for particles of vicinal cells. If the side of the cells is much smaller than the side of the complete system, the number of computations is drastically reduced.
"""

# ╔═╡ 5be87c6f-5c31-4d14-a8cb-4e63ef39d538
begin
    
function cell_list_picture()
    
    function square(c,side)
          x = [ c[1]-side/2, c[1]+side/2, c[1]+side/2, c[1]-side/2, c[1]-side/2]  
          y = [ c[2]-side/2, c[2]-side/2, c[2]+side/2, c[2]+side/2, c[2]-side/2]
          return x, y
    end
    
    plt = plot()
    
    x,y=square([5,5],2)
    plot!(
        plt,x,y,seriestype=[:shape],
        linewidth=2,fillalpha=0.05,color="green",label=""
    )
    
    x,y=square([5,5],6)
    plot!(
        plt,x,y,seriestype=[:shape],
          linewidth=2,fillalpha=0.05,color="orange",label=""
    )
    
    lines = collect(2:2:8)
    vline!(plt,lines,color="gray",label="",style=:dash)
    hline!(plt,lines,color="gray",label="",style=:dash)
    
    px = [ 0.1 + 9.8*rand() for i in 1:100 ]
    py = [ 0.1 + 9.8*rand() for i in 1:100 ]
    scatter!(plt,px,py,label="",alpha=0.20,color="blue")
    
    fontsize=8
    annotate!(plt,3,3,text("(i-1,j-1)",fontsize,:Courier))
    annotate!(plt,5,3,text("(i-1,j)",fontsize,:Courier))
    annotate!(plt,7,3,text("(i-1,j+1)",fontsize,:Courier))
    
    annotate!(plt,3,5,text("(i,j-1)",fontsize,:Courier))
    annotate!(plt,5,5,text("(i,j)",fontsize,:Courier))
    annotate!(plt,7,5,text("(i,j+1)",fontsize,:Courier))
    
    annotate!(plt,3,7,text("(i+1,j-1)",fontsize,:Courier))
    annotate!(plt,5,7,text("(i+1,j)",fontsize,:Courier))
    annotate!(plt,7,7,text("(i+1,j+1)",fontsize,:Courier))
    
    plot!(
        plt,size=(400,400), 
        xlim=(1.3,8.7),xticks=:none,
        ylim=(1.3,8.7),yticks=:none,
        framestyle=:box,
        xlabel="x",ylabel="y",grid=false
    )
    
    return plt
end

cell_list_picture()
end

# ╔═╡ 0c07edd3-c0a1-4f72-a16a-74badb7a6123
md"""
To use `CellListMap.jl` we need to setup our system, by providing the data on the box properties and the cutoff of the interactions:
"""

# ╔═╡ 4fc5ef4d-e072-41f7-aef9-b42730c8313c
box = Box([box_side,box_side],cutoff)

# ╔═╡ 19c5cc9d-8304-4e36-a3ea-a1151f28f71d
md"""
The particles are then classified in the cells. Virtual (ghost) particles are created at the boundaries to handle peridic boundary conditions and avoid having to wrap coordinates during the pairwise computation stage:
"""

# ╔═╡ 7dcadd85-2986-4e42-aa84-67128a8f666d
cl = CellList(x0_large,box)

# ╔═╡ 0b5c6ede-bceb-499a-a9a8-3c6a75ed340a
md"""
We only need to implement the function that has to be evaluated *if$ the particles are closer than the cutoff. This function will only be called in that case. Here, the function will update the force vector:
"""

# ╔═╡ 91b5eac1-4799-4a72-ac6a-e2b117b787d5
function fpair_cl(x,y,i,j,d2,f,box::Box)
    Δv = y - x
    d = sqrt(d2)
    fₓ = 2*(d - box.cutoff)*(Δv/d)
    f[i] += fₓ
    f[j] -= fₓ
    return f
end

# ╔═╡ 0f86ab3c-29aa-472b-8194-228c736ee940
md"""
The function that computes the forces in our simulation will, then, consist of an update of the cell lists followed by a call to the `map_pairwise!` function of `CellListMap.jl`, which takes as arguments the function to be mapped (`fpair_cl` here), the initial value of the forces vector `f`, and the system properties. We run only the serial version in this example:
"""

# ╔═╡ 0b8a2292-c0d6-44e4-b560-32d9d579a008
function forces_cl!(f::Vector{T},x,box::Box,cl::CellList,fpair::F) where {T,F}
    fill!(f,zero(T))
    cl = UpdateCellList!(x,box,cl,parallel=false)
    map_pairwise!(
        (x,y,i,j,d2,f) -> fpair(x,y,i,j,d2,f,box),
        f, box, cl, parallel=false
    )
    return f
end

# ╔═╡ d6585cca-78bf-41d1-aea3-01d9831d76cb
md"""
With a proper definition of the function to compute forces, we can now run again the simulation:
"""

# ╔═╡ 1b7b7d48-79d2-4317-9045-5b7e7bd073e5
t_cell_lists = @elapsed trajectory_cell_lists = md((
    x0 = x0_large, 
    v0 = [random_vec(Vec2D{Float64},(-1,1)) for _ in 1:n_large ], 
    mass = [ 10.0 for _ in 1:n_large ],
    dt = 0.1,
    nsteps = 1000,
    isave = 10,
    forces! = (f,x) -> forces_cl!(f,x,box,cl,fpair_cl)
)...)

# ╔═╡ 3f9dad58-294c-405c-bfc4-67855bb1e825
md""" 
Running time of CellListMap: $t_cell_lists seconds (on the second run - compilation takes about 2 seconds).
"""

# ╔═╡ 6d61b58f-b88f-48f4-8bdd-0bb1a8bc1c82
md"""
Even for a small system like this one, the speedup is significant (of about $(round(Int,t_naive/t_cell_lists)) times here). 
"""

# ╔═╡ 76b8695e-64dc-44bc-8938-ce22c4a9e4d0
md"""
## Energy minimization: the Packmol strategy
"""

# ╔═╡ 372637ff-9305-4d45-bf6e-e6531dadbd14
md"""
### A Lennard-Jones potential energy

Molecular dynamics simulations usually involve computing, for each pair of atoms, a Lennard-Jones function of the form:

$$u(r) = \varepsilon\left(\frac{\sigma^{12}}{r^{12}} - 2\frac{\sigma^6}{r^6}\right)$$

The high powers make the numerical behavior of this function quite undesirable. Let us try to minimize the energy of a randomly generated set of positions.

We will define one function that adds to the energy the contribution of a given pair of particles, and then use the `map_pairwise!` function of `CellListMap.jl` to compute this function for all pairs closer than a cutoff.
"""

# ╔═╡ 3b2c08a6-b27e-49be-a0b0-e5cb3d5546e0
md"""
The `FastPow` package unrols the high powers that need to be computed into multiplications, squares and cubes, which are faster to compute (with some loss of precision which is of no concern here). This could be done by hand, but for code clarity and convenience, we opt to use the `@fastpow` macro.
"""

# ╔═╡ 7280e368-c68a-48a5-91fe-93c76607c144
md"""
The function that computes the energy associated to one pair of particles is, then:
"""

# ╔═╡ 755fae26-6db9-45a0-a60d-d0e9c063f8aa
function ulj_pair(r2,u,ε,σ)
    @fastpow u += ε*(σ^12/r2^6 - 2*σ^6/r2^3)
    return u
end

# ╔═╡ 9a8d8012-ba54-4d9b-8c4c-fe6358508f2a
md"""
The function that computes the total energy is the mapping of that function to all relevant pairs through the `map_pairwise!` function:
"""

# ╔═╡ ffbeae5f-8aec-4473-a446-5b73bd911733
function ulj(x,ε,σ,box::Box,cl::CellList)
    cl = UpdateCellList!(x,box,cl,parallel=false)
    u = map_pairwise!(
        (x,y,i,j,d2,u) -> ulj_pair(d2,u,ε,σ),
        zero(eltype(σ)), box, cl,
        parallel=false
    )
    return u
end

# ╔═╡ 3738e40f-9596-469d-aa58-a4b28d8a22f8
md"""
The corresponding functions that updates the forces are:
"""

# ╔═╡ 5f1054b8-2337-43c1-a086-26233e95d42b
function flj_pair!(x,y,i,j,r2,f,ε,σ)
    @fastpow ∂u∂x = 12*ε*(σ^12/r2^7 - σ^6/r2^4)*(y-x)
    f[i] -= ∂u∂x
    f[j] += ∂u∂x
    return f
end

# ╔═╡ bd719619-bdd4-4c3c-8d66-1df0f210c595
function flj!(f::Vector{T},x,ε,σ,box,cl) where T
    cl = UpdateCellList!(x,box,cl,parallel=false)
    fill!(f,zero(T))
    map_pairwise!(
        (x,y,i,j,d2,f) -> flj_pair!(x,y,i,j,d2,f,ε,σ),
        f, box, cl, 
        parallel=false
    )
    return f
end

# ╔═╡ f289955b-0239-4b8d-ba08-2edf0a7284c2
md"""
### A physical system: Neon gas

Let us study a system of particles with an actual physical meaning, which is more interesting that a random set of points. 

We will compute the energy of a Ne gas with 10k particles, with density $\sim 0.1$ particles/ų, which is roughly the atomic density of liquid water. 

The Lennard-Jones parameters for Neon are:
"""

# ╔═╡ 878ab5f7-28c1-4832-9c58-cb36b360766f
const ε = 0.0441795 # kcal/mol

# ╔═╡ a0dcd888-059d-4abe-bb6b-958d2879101c
const σ = 2*1.64009 # Å

# ╔═╡ b1c5b7e5-cfbf-4d93-bd79-2924d957ae14
md"""
We chose to simulate 10k particles, for which an atomic density typical of room-temperature liquids results in the following box side (thus, we are actually simulating a highly-compressed Neon gas, but this is more interesting because the number or pairwise interactions which have to be computed is greater for denser systems):
"""

# ╔═╡ cd1102ac-1500-4d79-be83-72ac9180c7ce
const n_Ne = 10_000

# ╔═╡ f21604f4-e4f7-4d43-b3d9-32f429df443e
const box_side_Ne = (10_000/0.1)^(1/3)

# ╔═╡ 10826a95-16f8-416d-b8c1-0ef3347c9b20
x0_Ne = [random_vec(Vec3D{Float64},(0,box_side_Ne)) for _ in 1:n_Ne ]

# ╔═╡ c46a4f97-78e4-42fd-82b3-4dc6ce99abac
md"""
Given the initial coordinates, box size, and a typical cutoff of 12Å, we can initialize the system box and cell lists:
"""

# ╔═╡ 7c433791-a653-4836-91e2-084355c01d90
const box_Ne = Box([box_side_Ne for _ in 1:3],12.)

# ╔═╡ 410b9da4-7848-4385-bffc-a3d9bd03cf19
const cl_Ne = CellList(x0_Ne,box_Ne)

# ╔═╡ c08fff28-520e-40af-951c-fe3c324f67e0
md"""
### First, let us try to minimize the energy
"""

# ╔═╡ eb0f9080-2523-4633-be21-3a2281a1629e
md"""
The first thing in a MD simulation is trying to remove bad contacts by energy minimization:
"""

# ╔═╡ e3cc3c77-71ad-4006-8e27-fabaa1ae9cfb
md"""
The obtained energy (after 500 steps of stepest descent) is:
"""

# ╔═╡ 739c9a8a-13a5-4a33-a441-f5bc6cb35e82
md"""
### Packing the atoms
"""

# ╔═╡ 4e059cb8-6dac-450d-9f46-b3e657d9c3cf
md"""
To pack the atoms the "cutoff" needs to be of the order of the atom radii, instead of the cutoff of the Lennard-Jones interactions. Thus, we redefine the cell lists with a cutoff of σ/2. Since only very short-ranged interactions have to be computed, and the function is well behaved, the optimization is fast:
"""

# ╔═╡ 5ff9c89a-d999-4af2-8e7e-fb99d4948c36
md"""
Let us initialize again the system, considering the smaller cutoff:
"""

# ╔═╡ 0f2f55f6-060b-475e-bef7-eaa99da4d99f
box_pack = Box([box_side_Ne for _ in 1:3],σ/2)

# ╔═╡ 415ad590-247b-4a5d-b21e-7af4d0c17493
cl_pack = CellList(x0_Ne,box_pack)

# ╔═╡ 53c2e16e-b7f5-4df2-96f4-08402b5f8979
md"""
We previously defined the short-range forces in the `forces_cl` function, but we didn't use the "energy" associated to it, which we will use now:
"""

# ╔═╡ 16cdbc18-e846-4d0a-b7e6-87f07c0c52d9
function u_pack(x,box::Box,cl::CellList)
    cl = UpdateCellList!(x,box,cl,parallel=false)
    u = map_pairwise!(
        (x,y,i,j,d2,u) -> begin
			u += (sqrt(d2) - box.cutoff)^2 # objective function
			return u
		end,
        0., box, cl,
        parallel=false
    )
    return u
end

# ╔═╡ 79169f89-fedc-466b-8170-fff99b98e147
md"""
Given the packing energy and gradient, we can solve the packing problem:
"""

# ╔═╡ b5e258cd-5542-4a4c-ae0f-91c2fee426db
md"""
Importantly, the result is a packing function which converged to a global minimizer (the resulting packing function value is zero):
"""

# ╔═╡ b0dc1c2b-82b7-488d-8074-1ef9f59a15e5
md"""
The energy, on the other side, is not necessarily small:
"""

# ╔═╡ 97b8b15b-75c7-4321-999d-b067ed2a04f9
md"""
In two dimensions, this is the difference between a randomly generated set of coordinates, and a set obtained after solving the packing problem:
"""

# ╔═╡ 591e6a9c-444c-471f-a56b-4dfbc9111989
md"""
Even if the energy is high, we have the guarantee that no atoms are too close to each other, and this is an adequate configuration for a molecular dynamics simulation.

`Packmol` solves this packing problem for molecules of complex shape, allowing the user to specify different geometrical constraints that define the arrangements of the atoms in the system.
"""

# ╔═╡ 1f265576-824a-4764-a738-685554068079
md"""
## How fast is CellListMap.jl?

An idea of the efficiency of the cell list implementation in `CellListMap.jl` can be obtained by comparing the time required for an actual simulation of these Ne gas, compared to a stablished molecular dynamics simulation package, as [NAMD](https://www.ks.uiuc.edu/Research/namd/). 

We can run a simulation of this gas using the same functions we defined before, but with the actual potential energy:
"""

# ╔═╡ 1e099e1f-6494-419b-8517-5bded3e18aa6
md"""
# Remarks

1. Many types of distributions, for instance, of coordinates, can ben generated with the `Distributions.jl`  package. The generic character of the functions allow the functions to be used on custom types, as the `Vec2D` implemented here.

2. The integrator of our `md` function is only the simplest one. The standard integrator for MD simulations is `Velocity-Verlet`, and is implemented in the code that compares the performance of `CellListMap.jl` and NAMD. Many other integration algorithms are implemented, for example, in the `DifferentialEquations.jl` package.

3. The propagation of the uncertainty and the differentiability of particle simulations must be taken with a grain of salt. These systems are typically chaotic, thus uncertainties increase exponentialy, and derivatives are very unstable. An interesting blog post discussing the sensitivity of these calculations is [here](https://frankschae.github.io/post/shadowing/), and again the `DifferentialEquations.jl` package provides more adequate tools to deal with parameter optimization under such circunstances. 

"""

# ╔═╡ 10c86547-d4f4-4c3f-8906-ac18ce93f3b6
md"""
# Acknowledgements

The author thanks [Mosè Giordano](https://giordano.github.io/) for valuable discussions on the working of `Measurements`, and many other members of the Julia and Fortran discourse forums for indirect contributions to this work. We also thank the FortranCon organizing comitee, in particular [Milan Curcic](https://milancurcic.com/) and [Ondřej Čertík](https://github.com/certik) for the kind invitation and great contributions for the developement of a modern community and tools around Fortran.
"""

# ╔═╡ 2871aca3-e6b4-4a2d-868a-36562e9a274c
md"""
# Some notebook options and setup
"""

# ╔═╡ 2a2e9155-1c77-46fd-8502-8431573f94d0
md"""
## Default plot setup
"""

# ╔═╡ b557a646-8c3c-4dc7-8788-bf98aec8c5c0
md"""
Use Printf to print some data.
"""

# ╔═╡ 260d5753-6cc2-4137-8a2c-8d8a47585ecf
md"""
We can set this to false to avoid ploting everything. 
"""

# ╔═╡ a9981931-4cc9-4d16-a6d2-34b4071a84d7
const build_plots = true

# ╔═╡ 374f239b-6470-40ed-b068-a8ecaace4f09
build_plots && begin
    r = 0:0.1:1.2*cutoff
    plot(layout=(1,2),size=(600,300))
    plot!(r,energy.(0.,r,cutoff),xlabel="Distance",ylabel="Energy",subplot=1)
    plot!(r,fₓ.(0.,r,cutoff),xlabel="Distance",ylabel="Force",subplot=2)
end    

# ╔═╡ 43e6b146-ee35-40f1-b540-3da22b9e1b1b
build_plots && scatter([(x.x, x.y) for x in x0])

# ╔═╡ 505ef5ab-f131-4ab3-a723-795b5eb5dc0f
build_plots && @gif for (step,x) in pairs(trajectory)
    scatter([ (p.x,p.y) for p in x ], lims=(-250,250))
    annotate!(130,-210,text("step: $step",plot_font,12,:left))
end

# ╔═╡ efc586a2-0946-4dc5-ab3a-3902a811f3ad
build_plots && @gif for (step,x) in pairs(trajectory_periodic)
    scatter([ wrap.((p.x,p.y),100) for p in x ], lims=(-60,60))
    annotate!(25,-50,text("step: $step",plot_font,12,:left))
end

# ╔═╡ 4a498c18-406f-4437-b378-aa9fdc75b919
build_plots && @gif for x in trajectory_periodic_3D
    scatter([ wrap.((p.x,p.y,p.z),100) for p in x ], lims=(-60,60))
end

# ╔═╡ d87c22d1-d595-4d43-ab1c-f28d282a3485
build_plots && ( trajectory_2D_error = md((
    x0 = [random_vec(Vec2D{Measurement{Float64}},(-50,50),1e-5) for _ in 1:100 ], 
    v0 = [random_vec(Vec2D{Measurement{Float64}},(-1,1),1e-5) for _ in 1:100 ],
    mass = [ 1.0 for _ in 1:100 ],
    dt = 0.1,
    nsteps = 100,
    isave = 1,
    forces! = (f,x) -> forces!(f,x, (i,j,p1,p2) -> fₓ(p1,p2,cutoff,side))
)...) )

# ╔═╡ bf0a5303-f5ce-4711-b9ee-a12ce2d8a397
build_plots && @gif for x in trajectory_2D_error
    positions = [ wrap.((p.x.val,p.y.val),100) for p in x ]
    scatter(positions, lims=(-60,60))
end

# ╔═╡ e24ce081-e367-4feb-8a79-66b8654a0b3a
build_plots && @gif for x in trajectory_2D_error
    histogram(
        [ p.x.err for p in x ],
        xlabel="Uncertainty in x",ylabel="Number of points",
        bins=0:1e-4:20e-4,ylims=[0,50]
    )
end

# ╔═╡ 1067527e-76b7-4331-b3ab-efd72fb99dfc
build_plots && @gif for (step,x) in pairs(trajectory_planets)
    colors = [ :yellow, :grey, :brown, :blue, :red ]
    positions = [ (p.x.val,p.y.val) for p in x ]
    xerr = [ p.x.err for p in x ]
    yerr = [ p.y.err for p in x ] 
    scatter(positions,lims=[-250,250], markercolor=colors, xerror=xerr, yerror=yerr)
    annotate!(150,-210,text(@sprintf("%5i days",step),plot_font,12))
end

# ╔═╡ 4cef9cea-1e84-42b9-bff6-b9a8b3bfe8da
build_plots && @gif for step in eachindex(earth_traj_best)
    colors = [ :yellow, :blue ]
    positions0 = [ (p.x,p.y) for p in earth_traj_0[step] ] 
    positions_best = [ (p.x,p.y) for p in earth_traj_best[step] ]
    scatter(positions0,lims=[-250,250], markercolor=colors, alpha=0.5)
    scatter!(positions_best,lims=[-250,250], markercolor=colors)
    scatter!(
        (earth_traj_best[1][2].x,earth_traj_best[1][2].y),
        markercolor=:white,alpha=0.5,
        markersize=10
    )
    annotate!(150,-210,text(@sprintf("%5i days",2*step),plot_font,12))
end

# ╔═╡ e5b557d7-0952-4409-ae4c-a0c8ce736e03
build_plots && @gif for (step,x) in pairs(trajectory_periodic_large)
    scatter(
        [ wrap.((p.x,p.y),box_side) for p in x ], 
        lims=(-1.1*box_side/2,1.1*box_side/2)
    )
end

# ╔═╡ 30d2f39e-5df2-4f38-8032-e5f8492ba335
build_plots && @gif for (step,x) in pairs(trajectory_cell_lists)
    scatter(
        [ wrap.((p.x,p.y),box_side) for p in x ], 
        lims=(-1.1*box_side/2,1.1*box_side/2)
    )
end

# ╔═╡ 2634feff-7442-4d8f-b8e5-c11113136980
build_plots && begin
	r_lj = 2.8:0.05:8
	plot(
		r_lj,ulj_pair.(r_lj.^2,0.,ε,σ),
		xlabel="Distance / Å", ylabel="Potential energy / kcal/mol"
	)
end

# ╔═╡ b4154fb7-e0b0-4211-8490-8a8fe47cd2da
md"""
## Gradient descent for vectors
"""

# ╔═╡ 8ac7b1bf-c958-4eb5-8376-f802b372e796
function gradient_descent!(x::Vector{T},f,g!;tol=1e-3,maxtrial=500) where T
    gnorm(x) = maximum(norm(v) for v in x)
    itrial = 0
    step = 1.0
    xtrial = similar(x)
    g = fill!(similar(x),zero(T))
    fx = f(x)
    g = g!(g,x)
    while (gnorm(g) > tol) && (itrial < maxtrial) 
        @. xtrial = x - step*g
        ftrial = f(xtrial)  
        if ftrial >= fx
            step = step / 2
        else
            x .= xtrial
            fx = ftrial
            g = g!(g,x)
            step = step * 2
        end
        @show itrial, step, fx, ftrial, gnorm(g)
		itrial += 1
    end 
    return x
end

# ╔═╡ 357c6621-b2b8-4f30-ba41-ffc1ae6f031b
t_min = @elapsed x_min = gradient_descent!(
    copy(x0_Ne),
    (x) -> ulj(x,ε,σ,box_Ne,cl_Ne),
    (g,x) -> -flj!(g,x,ε,σ,box_Ne,cl_Ne)
)

# ╔═╡ 58eb5b4b-76ad-4f7a-b86b-0494a857dca1
ulj(x_min,ε,σ,box_Ne,cl_Ne)

# ╔═╡ 574047fa-6626-4cd0-8317-32118129711e
md"""
Here the example is only illustrative, but shows a common behavior: the energy after minimization is still too high. The cost and numerical instability of the true potential, at short distances, make it hard to minimize.

**Time required for energy minimization: $t_min seconds**

Because of that [`Packmol`](http://m3g.iqm.unicamp.br/packmol) was introduced. We first solve the problem of packing the atoms in the space guaranteeing a minimum distance between the atoms. Here, this consists on the minimization of our simplified potential:
"""

# ╔═╡ 224336e2-522c-44af-b9a1-307e2ffff0f9
t_pack = @elapsed x_pack = gradient_descent!(
    copy(x0_Ne),
    (x) -> u_pack(x,box_pack,cl_pack),
    (g,x) -> -forces_cl!(g,x,box_pack,cl_pack,fpair_cl)
)

# ╔═╡ 06526edf-911a-4ecc-a350-6d932ca56cd5
md"""
**Time required for packing: $t_pack seconds**
"""

# ╔═╡ c48210e0-1a04-4f84-a4e2-f6b5d34a603d
u_pack(x_pack,box_pack,cl_pack)

# ╔═╡ 0471b987-f987-4656-b961-285e32d0a5e1
ulj(x_pack,ε,σ,box_Ne,cl_Ne)

# ╔═╡ 339487cd-8ee8-4d1d-984b-b4c5ff00bae3
t_Ne = @elapsed trajectory_Ne = md((
    x0 = x_pack, 
    v0 = [random_vec(Vec3D{Float64},(-1,1)) for _ in 1:n_Ne ], 
    mass = [ 20.179 for _ in 1:n_Ne ],
    dt = 0.01,
    nsteps = 100,
    isave = 10,
    forces! = (f,x) -> flj!(f,x,ε,σ,box_Ne,cl_Ne)
)...)

# ╔═╡ 9cb29b01-7f49-4145-96d8-c8fd971fe1c8
build_plots && @gif for x in trajectory_Ne
    scatter(
      [ wrap.((p.x,p.y,p.z),box_side_Ne) for p in x ], 
      lims=(-1.1*box_side_Ne/2,1.1*box_side_Ne/2)
    )
end

# ╔═╡ ac52a71b-1138-4f1b-99c3-c174d9f09187
md"""
This simulation took $t_Ne seconds. 

Again, to understand exactly what that means, we need to perform a proper comparison. In [this benchmark](https://github.com/m3g/2021_FortranCon/tree/main/celllistmap_vs_namd) two simulations of the same gas, with proper thermodynamic conditions, are performed. The Julia algorithm implemented is similar to the present one, except that thermalization is done by velocity rescaling and the velocity-verlet algorithm is used for propagating the positions. The same algorthms are used in the equivalent NAMD simulations. The benchmark result is, for a 4-cores/8-threads execution in my Laptop:

````
NAMD:

real    1m14,049s
user    8m59,065s
sys     0m1,130s

CellListMap:

real    1m21,054s
user    7m38,053s
sys     0m2,172s
````

This comparison of course can be questined: `NAMD` is a general purpose MD package designed for massive-parallel simulations, and `CellListMap.jl` is a package for computing any distance-dependent property, and for now designed and optimized on shared-memory computers. 

Nevertheless,the benchmark shows that it is possible to write high-performant code in Julia, and that `CellListMap.jl` is a powerful tool for simulating or computing distance-dependent properties from the results of simulations.

One of the applications of this package is the computation of distribution functions, in the [ComplexMixtures.jl](https://m3g.github.io/ComplexMixtures.jl/stable/)  package.

"""

# ╔═╡ b5008faf-fd43-45dd-a5a1-7f51e0b4ede5
md"""
## Table of Contents
"""

# ╔═╡ 555d1f62-b95b-4377-a8e2-9e442ee7526d


# ╔═╡ f5510c1e-9b9f-49f0-bc7e-0fd8e79a5760
md"""
## 2D packing example
"""

# ╔═╡ 5d9a40b5-4050-47d2-9855-e9b62d56e8df
side_test = 50

# ╔═╡ 7f556f7c-cdb0-4f91-a359-2f933bbc5b68
xtest = [ random_vec(Vec2D{Float64},(-side_test,side_test)) for _ in 1:1000 ]

# ╔═╡ 0fc843d2-ac4f-4717-a298-92a476223112
tol_test = 2

# ╔═╡ fc7f665b-00d9-431b-a97e-d2ff7253221a
box_test = Box([side_test,side_test],tol_test)

# ╔═╡ aadf6e48-0cbf-4973-86a1-173b6648d1df
cl_test = CellList(xtest,box_test)

# ╔═╡ 452e1ea7-98be-4910-ba6b-c0881fb251b2
x_pack_test = gradient_descent!(
    copy(xtest),
    (x) -> u_pack(x,box_test,cl_test),
    (g,x) -> -forces_cl!(g,x,box_test,cl_test,fpair_cl)
)

# ╔═╡ d9f254dc-ae4a-40b3-b682-f8a501e10a2d
build_plots && begin
	plot(layout=(1,2))
	scatter!([ wrap.(Tuple(p),side_test) for p in xtest ],subplot=1)
	scatter!([ wrap.(Tuple(p),side_test) for p in x_pack_test ],subplot=2)
	plot!(lims=(-1.1*side_test/2,1.1*side_test/2),aspect_ratio=1,size=(800,400))
end

# ╔═╡ 00000000-0000-0000-0000-000000000001
PLUTO_PROJECT_TOML_CONTENTS = """
[deps]
BenchmarkTools = "6e4b80f9-dd63-53aa-95a3-0cdb28fa8baf"
CellListMap = "69e1c6dd-3888-40e6-b3c8-31ac5f578864"
FastPow = "c0e83750-1142-43a8-81cf-6c956b72b4d1"
ForwardDiff = "f6369f11-7733-5829-9624-2563aa707210"
LinearAlgebra = "37e2e46d-f89d-539d-b4ee-838fcccc9c8e"
Measurements = "eff96d63-e80a-5855-80a2-b1b0885c5ab7"
Plots = "91a5bcdd-55d7-5caf-9e0b-520d859cae80"
PlutoUI = "7f904dfe-b85e-4ff6-b463-dae2292396a8"
Printf = "de0858da-6303-5e67-8744-51eddeeeb8d7"
StaticArrays = "90137ffa-7385-5640-81b9-e52037218182"

[compat]
BenchmarkTools = "~1.2.0"
CellListMap = "~0.5.20"
FastPow = "~0.1.0"
ForwardDiff = "~0.10.19"
Measurements = "~2.6.0"
Plots = "~1.22.2"
PlutoUI = "~0.7.11"
StaticArrays = "~1.2.12"
"""

# ╔═╡ 00000000-0000-0000-0000-000000000002
PLUTO_MANIFEST_TOML_CONTENTS = """
# This file is machine-generated - editing it directly is not advised

[[Adapt]]
deps = ["LinearAlgebra"]
git-tree-sha1 = "84918055d15b3114ede17ac6a7182f68870c16f7"
uuid = "79e6a3ab-5dfb-504d-930d-738a2a938a0e"
version = "3.3.1"

[[ArgTools]]
uuid = "0dad84c5-d112-42e6-8d28-ef12dabb789f"

[[Artifacts]]
uuid = "56f22d72-fd6d-98f1-02f0-08ddc0907c33"

[[Base64]]
uuid = "2a0f44e3-6c83-55bd-87e4-b1978d98bd5f"

[[BenchmarkTools]]
deps = ["JSON", "Logging", "Printf", "Profile", "Statistics", "UUIDs"]
git-tree-sha1 = "61adeb0823084487000600ef8b1c00cc2474cd47"
uuid = "6e4b80f9-dd63-53aa-95a3-0cdb28fa8baf"
version = "1.2.0"

[[Bzip2_jll]]
deps = ["Artifacts", "JLLWrappers", "Libdl", "Pkg"]
git-tree-sha1 = "19a35467a82e236ff51bc17a3a44b69ef35185a2"
uuid = "6e34b625-4abd-537c-b88f-471c36dfa7a0"
version = "1.0.8+0"

[[Cairo_jll]]
deps = ["Artifacts", "Bzip2_jll", "Fontconfig_jll", "FreeType2_jll", "Glib_jll", "JLLWrappers", "LZO_jll", "Libdl", "Pixman_jll", "Pkg", "Xorg_libXext_jll", "Xorg_libXrender_jll", "Zlib_jll", "libpng_jll"]
git-tree-sha1 = "f2202b55d816427cd385a9a4f3ffb226bee80f99"
uuid = "83423d85-b0ee-5818-9007-b63ccbeb887a"
version = "1.16.1+0"

[[Calculus]]
deps = ["LinearAlgebra"]
git-tree-sha1 = "f641eb0a4f00c343bbc32346e1217b86f3ce9dad"
uuid = "49dc2e85-a5d0-5ad3-a950-438e2897f1b9"
version = "0.5.1"

[[CellListMap]]
deps = ["DocStringExtensions", "LinearAlgebra", "Parameters", "ProgressMeter", "Random", "Setfield", "StaticArrays"]
git-tree-sha1 = "6307da1063024d02438f460b98a7539fc6270298"
uuid = "69e1c6dd-3888-40e6-b3c8-31ac5f578864"
version = "0.5.20"

[[ChainRulesCore]]
deps = ["Compat", "LinearAlgebra", "SparseArrays"]
git-tree-sha1 = "bd4afa1fdeec0c8b89dad3c6e92bc6e3b0fec9ce"
uuid = "d360d2e6-b24c-11e9-a2a3-2a2ae2dbcce4"
version = "1.6.0"

[[ColorSchemes]]
deps = ["ColorTypes", "Colors", "FixedPointNumbers", "Random"]
git-tree-sha1 = "a851fec56cb73cfdf43762999ec72eff5b86882a"
uuid = "35d6a980-a343-548e-a6ea-1d62b119f2f4"
version = "3.15.0"

[[ColorTypes]]
deps = ["FixedPointNumbers", "Random"]
git-tree-sha1 = "024fe24d83e4a5bf5fc80501a314ce0d1aa35597"
uuid = "3da002f7-5984-5a60-b8a6-cbb66c0b333f"
version = "0.11.0"

[[Colors]]
deps = ["ColorTypes", "FixedPointNumbers", "Reexport"]
git-tree-sha1 = "417b0ed7b8b838aa6ca0a87aadf1bb9eb111ce40"
uuid = "5ae59095-9a9b-59fe-a467-6f913c188581"
version = "0.12.8"

[[CommonSubexpressions]]
deps = ["MacroTools", "Test"]
git-tree-sha1 = "7b8a93dba8af7e3b42fecabf646260105ac373f7"
uuid = "bbf7d656-a473-5ed7-a52c-81e309532950"
version = "0.3.0"

[[Compat]]
deps = ["Base64", "Dates", "DelimitedFiles", "Distributed", "InteractiveUtils", "LibGit2", "Libdl", "LinearAlgebra", "Markdown", "Mmap", "Pkg", "Printf", "REPL", "Random", "SHA", "Serialization", "SharedArrays", "Sockets", "SparseArrays", "Statistics", "Test", "UUIDs", "Unicode"]
git-tree-sha1 = "31d0151f5716b655421d9d75b7fa74cc4e744df2"
uuid = "34da2185-b29b-5c13-b0c7-acf172513d20"
version = "3.39.0"

[[CompilerSupportLibraries_jll]]
deps = ["Artifacts", "Libdl"]
uuid = "e66e0078-7015-5450-92f7-15fbd957f2ae"

[[ConstructionBase]]
deps = ["LinearAlgebra"]
git-tree-sha1 = "f74e9d5388b8620b4cee35d4c5a618dd4dc547f4"
uuid = "187b0558-2788-49d3-abe0-74a17ed4e7c9"
version = "1.3.0"

[[Contour]]
deps = ["StaticArrays"]
git-tree-sha1 = "9f02045d934dc030edad45944ea80dbd1f0ebea7"
uuid = "d38c429a-6771-53c6-b99e-75d170b6e991"
version = "0.5.7"

[[DataAPI]]
git-tree-sha1 = "cc70b17275652eb47bc9e5f81635981f13cea5c8"
uuid = "9a962f9c-6df0-11e9-0e5d-c546b8b5ee8a"
version = "1.9.0"

[[DataStructures]]
deps = ["Compat", "InteractiveUtils", "OrderedCollections"]
git-tree-sha1 = "7d9d316f04214f7efdbb6398d545446e246eff02"
uuid = "864edb3b-99cc-5e75-8d2d-829cb0a9cfe8"
version = "0.18.10"

[[DataValueInterfaces]]
git-tree-sha1 = "bfc1187b79289637fa0ef6d4436ebdfe6905cbd6"
uuid = "e2d170a0-9d28-54be-80f0-106bbe20a464"
version = "1.0.0"

[[Dates]]
deps = ["Printf"]
uuid = "ade2ca70-3891-5945-98fb-dc099432e06a"

[[DelimitedFiles]]
deps = ["Mmap"]
uuid = "8bb1440f-4735-579b-a4ab-409b98df4dab"

[[DiffResults]]
deps = ["StaticArrays"]
git-tree-sha1 = "c18e98cba888c6c25d1c3b048e4b3380ca956805"
uuid = "163ba53b-c6d8-5494-b064-1a9d43ac40c5"
version = "1.0.3"

[[DiffRules]]
deps = ["NaNMath", "Random", "SpecialFunctions"]
git-tree-sha1 = "7220bc21c33e990c14f4a9a319b1d242ebc5b269"
uuid = "b552c78f-8df3-52c6-915a-8e097449b14b"
version = "1.3.1"

[[Distributed]]
deps = ["Random", "Serialization", "Sockets"]
uuid = "8ba89e20-285c-5b6f-9357-94700520ee1b"

[[DocStringExtensions]]
deps = ["LibGit2"]
git-tree-sha1 = "a32185f5428d3986f47c2ab78b1f216d5e6cc96f"
uuid = "ffbed154-4ef7-542d-bbb7-c09d3a79fcae"
version = "0.8.5"

[[Downloads]]
deps = ["ArgTools", "LibCURL", "NetworkOptions"]
uuid = "f43a241f-c20a-4ad4-852c-f6b1247861c6"

[[EarCut_jll]]
deps = ["Artifacts", "JLLWrappers", "Libdl", "Pkg"]
git-tree-sha1 = "3f3a2501fa7236e9b911e0f7a588c657e822bb6d"
uuid = "5ae413db-bbd1-5e63-b57d-d24a61df00f5"
version = "2.2.3+0"

[[Expat_jll]]
deps = ["Artifacts", "JLLWrappers", "Libdl", "Pkg"]
git-tree-sha1 = "b3bfd02e98aedfa5cf885665493c5598c350cd2f"
uuid = "2e619515-83b5-522b-bb60-26c02a35a201"
version = "2.2.10+0"

[[FFMPEG]]
deps = ["FFMPEG_jll"]
git-tree-sha1 = "b57e3acbe22f8484b4b5ff66a7499717fe1a9cc8"
uuid = "c87230d0-a227-11e9-1b43-d7ebe4e7570a"
version = "0.4.1"

[[FFMPEG_jll]]
deps = ["Artifacts", "Bzip2_jll", "FreeType2_jll", "FriBidi_jll", "JLLWrappers", "LAME_jll", "Libdl", "Ogg_jll", "OpenSSL_jll", "Opus_jll", "Pkg", "Zlib_jll", "libass_jll", "libfdk_aac_jll", "libvorbis_jll", "x264_jll", "x265_jll"]
git-tree-sha1 = "d8a578692e3077ac998b50c0217dfd67f21d1e5f"
uuid = "b22a6f82-2f65-5046-a5b2-351ab43fb4e5"
version = "4.4.0+0"

[[FastPow]]
git-tree-sha1 = "7d961335144dad74de0e1b3a9b60e4d114a78dc2"
uuid = "c0e83750-1142-43a8-81cf-6c956b72b4d1"
version = "0.1.0"

[[FixedPointNumbers]]
deps = ["Statistics"]
git-tree-sha1 = "335bfdceacc84c5cdf16aadc768aa5ddfc5383cc"
uuid = "53c48c17-4a7d-5ca2-90c5-79b7896eea93"
version = "0.8.4"

[[Fontconfig_jll]]
deps = ["Artifacts", "Bzip2_jll", "Expat_jll", "FreeType2_jll", "JLLWrappers", "Libdl", "Libuuid_jll", "Pkg", "Zlib_jll"]
git-tree-sha1 = "21efd19106a55620a188615da6d3d06cd7f6ee03"
uuid = "a3f928ae-7b40-5064-980b-68af3947d34b"
version = "2.13.93+0"

[[Formatting]]
deps = ["Printf"]
git-tree-sha1 = "8339d61043228fdd3eb658d86c926cb282ae72a8"
uuid = "59287772-0a20-5a39-b81b-1366585eb4c0"
version = "0.4.2"

[[ForwardDiff]]
deps = ["CommonSubexpressions", "DiffResults", "DiffRules", "LinearAlgebra", "NaNMath", "Printf", "Random", "SpecialFunctions", "StaticArrays"]
git-tree-sha1 = "b5e930ac60b613ef3406da6d4f42c35d8dc51419"
uuid = "f6369f11-7733-5829-9624-2563aa707210"
version = "0.10.19"

[[FreeType2_jll]]
deps = ["Artifacts", "Bzip2_jll", "JLLWrappers", "Libdl", "Pkg", "Zlib_jll"]
git-tree-sha1 = "87eb71354d8ec1a96d4a7636bd57a7347dde3ef9"
uuid = "d7e528f0-a631-5988-bf34-fe36492bcfd7"
version = "2.10.4+0"

[[FriBidi_jll]]
deps = ["Artifacts", "JLLWrappers", "Libdl", "Pkg"]
git-tree-sha1 = "aa31987c2ba8704e23c6c8ba8a4f769d5d7e4f91"
uuid = "559328eb-81f9-559d-9380-de523a88c83c"
version = "1.0.10+0"

[[Future]]
deps = ["Random"]
uuid = "9fa8497b-333b-5362-9e8d-4d0656e87820"

[[GLFW_jll]]
deps = ["Artifacts", "JLLWrappers", "Libdl", "Libglvnd_jll", "Pkg", "Xorg_libXcursor_jll", "Xorg_libXi_jll", "Xorg_libXinerama_jll", "Xorg_libXrandr_jll"]
git-tree-sha1 = "dba1e8614e98949abfa60480b13653813d8f0157"
uuid = "0656b61e-2033-5cc2-a64a-77c0f6c09b89"
version = "3.3.5+0"

[[GR]]
deps = ["Base64", "DelimitedFiles", "GR_jll", "HTTP", "JSON", "Libdl", "LinearAlgebra", "Pkg", "Printf", "Random", "Serialization", "Sockets", "Test", "UUIDs"]
git-tree-sha1 = "c2178cfbc0a5a552e16d097fae508f2024de61a3"
uuid = "28b8d3ca-fb5f-59d9-8090-bfdbd6d07a71"
version = "0.59.0"

[[GR_jll]]
deps = ["Artifacts", "Bzip2_jll", "Cairo_jll", "FFMPEG_jll", "Fontconfig_jll", "GLFW_jll", "JLLWrappers", "JpegTurbo_jll", "Libdl", "Libtiff_jll", "Pixman_jll", "Pkg", "Qt5Base_jll", "Zlib_jll", "libpng_jll"]
git-tree-sha1 = "ef49a187604f865f4708c90e3f431890724e9012"
uuid = "d2c73de3-f751-5644-a686-071e5b155ba9"
version = "0.59.0+0"

[[GeometryBasics]]
deps = ["EarCut_jll", "IterTools", "LinearAlgebra", "StaticArrays", "StructArrays", "Tables"]
git-tree-sha1 = "58bcdf5ebc057b085e58d95c138725628dd7453c"
uuid = "5c1252a2-5f33-56bf-86c9-59e7332b4326"
version = "0.4.1"

[[Gettext_jll]]
deps = ["Artifacts", "CompilerSupportLibraries_jll", "JLLWrappers", "Libdl", "Libiconv_jll", "Pkg", "XML2_jll"]
git-tree-sha1 = "9b02998aba7bf074d14de89f9d37ca24a1a0b046"
uuid = "78b55507-aeef-58d4-861c-77aaff3498b1"
version = "0.21.0+0"

[[Glib_jll]]
deps = ["Artifacts", "Gettext_jll", "JLLWrappers", "Libdl", "Libffi_jll", "Libiconv_jll", "Libmount_jll", "PCRE_jll", "Pkg", "Zlib_jll"]
git-tree-sha1 = "7bf67e9a481712b3dbe9cb3dac852dc4b1162e02"
uuid = "7746bdde-850d-59dc-9ae8-88ece973131d"
version = "2.68.3+0"

[[Graphite2_jll]]
deps = ["Artifacts", "JLLWrappers", "Libdl", "Pkg"]
git-tree-sha1 = "344bf40dcab1073aca04aa0df4fb092f920e4011"
uuid = "3b182d85-2403-5c21-9c21-1e1f0cc25472"
version = "1.3.14+0"

[[Grisu]]
git-tree-sha1 = "53bb909d1151e57e2484c3d1b53e19552b887fb2"
uuid = "42e2da0e-8278-4e71-bc24-59509adca0fe"
version = "1.0.2"

[[HTTP]]
deps = ["Base64", "Dates", "IniFile", "Logging", "MbedTLS", "NetworkOptions", "Sockets", "URIs"]
git-tree-sha1 = "60ed5f1643927479f845b0135bb369b031b541fa"
uuid = "cd3eb016-35fb-5094-929b-558a96fad6f3"
version = "0.9.14"

[[HarfBuzz_jll]]
deps = ["Artifacts", "Cairo_jll", "Fontconfig_jll", "FreeType2_jll", "Glib_jll", "Graphite2_jll", "JLLWrappers", "Libdl", "Libffi_jll", "Pkg"]
git-tree-sha1 = "8a954fed8ac097d5be04921d595f741115c1b2ad"
uuid = "2e76f6c2-a576-52d4-95c1-20adfe4de566"
version = "2.8.1+0"

[[HypertextLiteral]]
git-tree-sha1 = "72053798e1be56026b81d4e2682dbe58922e5ec9"
uuid = "ac1192a8-f4b3-4bfe-ba22-af5b92cd3ab2"
version = "0.9.0"

[[IOCapture]]
deps = ["Logging", "Random"]
git-tree-sha1 = "f7be53659ab06ddc986428d3a9dcc95f6fa6705a"
uuid = "b5f81e59-6552-4d32-b1f0-c071b021bf89"
version = "0.2.2"

[[IniFile]]
deps = ["Test"]
git-tree-sha1 = "098e4d2c533924c921f9f9847274f2ad89e018b8"
uuid = "83e8ac13-25f8-5344-8a64-a9f2b223428f"
version = "0.5.0"

[[InteractiveUtils]]
deps = ["Markdown"]
uuid = "b77e0a4c-d291-57a0-90e8-8db25a27a240"

[[IrrationalConstants]]
git-tree-sha1 = "f76424439413893a832026ca355fe273e93bce94"
uuid = "92d709cd-6900-40b7-9082-c6be49f344b6"
version = "0.1.0"

[[IterTools]]
git-tree-sha1 = "05110a2ab1fc5f932622ffea2a003221f4782c18"
uuid = "c8e1da08-722c-5040-9ed9-7db0dc04731e"
version = "1.3.0"

[[IteratorInterfaceExtensions]]
git-tree-sha1 = "a3f24677c21f5bbe9d2a714f95dcd58337fb2856"
uuid = "82899510-4779-5014-852e-03e436cf321d"
version = "1.0.0"

[[JLLWrappers]]
deps = ["Preferences"]
git-tree-sha1 = "642a199af8b68253517b80bd3bfd17eb4e84df6e"
uuid = "692b3bcd-3c85-4b1f-b108-f13ce0eb3210"
version = "1.3.0"

[[JSON]]
deps = ["Dates", "Mmap", "Parsers", "Unicode"]
git-tree-sha1 = "8076680b162ada2a031f707ac7b4953e30667a37"
uuid = "682c06a0-de6a-54ab-a142-c8b1cf79cde6"
version = "0.21.2"

[[JpegTurbo_jll]]
deps = ["Artifacts", "JLLWrappers", "Libdl", "Pkg"]
git-tree-sha1 = "d735490ac75c5cb9f1b00d8b5509c11984dc6943"
uuid = "aacddb02-875f-59d6-b918-886e6ef4fbf8"
version = "2.1.0+0"

[[LAME_jll]]
deps = ["Artifacts", "JLLWrappers", "Libdl", "Pkg"]
git-tree-sha1 = "f6250b16881adf048549549fba48b1161acdac8c"
uuid = "c1c5ebd0-6772-5130-a774-d5fcae4a789d"
version = "3.100.1+0"

[[LZO_jll]]
deps = ["Artifacts", "JLLWrappers", "Libdl", "Pkg"]
git-tree-sha1 = "e5b909bcf985c5e2605737d2ce278ed791b89be6"
uuid = "dd4b983a-f0e5-5f8d-a1b7-129d4a5fb1ac"
version = "2.10.1+0"

[[LaTeXStrings]]
git-tree-sha1 = "c7f1c695e06c01b95a67f0cd1d34994f3e7db104"
uuid = "b964fa9f-0449-5b57-a5c2-d3ea65f4040f"
version = "1.2.1"

[[Latexify]]
deps = ["Formatting", "InteractiveUtils", "LaTeXStrings", "MacroTools", "Markdown", "Printf", "Requires"]
git-tree-sha1 = "a4b12a1bd2ebade87891ab7e36fdbce582301a92"
uuid = "23fbe1c1-3f47-55db-b15f-69d7ec21a316"
version = "0.15.6"

[[LibCURL]]
deps = ["LibCURL_jll", "MozillaCACerts_jll"]
uuid = "b27032c2-a3e7-50c8-80cd-2d36dbcbfd21"

[[LibCURL_jll]]
deps = ["Artifacts", "LibSSH2_jll", "Libdl", "MbedTLS_jll", "Zlib_jll", "nghttp2_jll"]
uuid = "deac9b47-8bc7-5906-a0fe-35ac56dc84c0"

[[LibGit2]]
deps = ["Base64", "NetworkOptions", "Printf", "SHA"]
uuid = "76f85450-5226-5b5a-8eaa-529ad045b433"

[[LibSSH2_jll]]
deps = ["Artifacts", "Libdl", "MbedTLS_jll"]
uuid = "29816b5a-b9ab-546f-933c-edad1886dfa8"

[[Libdl]]
uuid = "8f399da3-3557-5675-b5ff-fb832c97cbdb"

[[Libffi_jll]]
deps = ["Artifacts", "JLLWrappers", "Libdl", "Pkg"]
git-tree-sha1 = "761a393aeccd6aa92ec3515e428c26bf99575b3b"
uuid = "e9f186c6-92d2-5b65-8a66-fee21dc1b490"
version = "3.2.2+0"

[[Libgcrypt_jll]]
deps = ["Artifacts", "JLLWrappers", "Libdl", "Libgpg_error_jll", "Pkg"]
git-tree-sha1 = "64613c82a59c120435c067c2b809fc61cf5166ae"
uuid = "d4300ac3-e22c-5743-9152-c294e39db1e4"
version = "1.8.7+0"

[[Libglvnd_jll]]
deps = ["Artifacts", "JLLWrappers", "Libdl", "Pkg", "Xorg_libX11_jll", "Xorg_libXext_jll"]
git-tree-sha1 = "7739f837d6447403596a75d19ed01fd08d6f56bf"
uuid = "7e76a0d4-f3c7-5321-8279-8d96eeed0f29"
version = "1.3.0+3"

[[Libgpg_error_jll]]
deps = ["Artifacts", "JLLWrappers", "Libdl", "Pkg"]
git-tree-sha1 = "c333716e46366857753e273ce6a69ee0945a6db9"
uuid = "7add5ba3-2f88-524e-9cd5-f83b8a55f7b8"
version = "1.42.0+0"

[[Libiconv_jll]]
deps = ["Artifacts", "JLLWrappers", "Libdl", "Pkg"]
git-tree-sha1 = "42b62845d70a619f063a7da093d995ec8e15e778"
uuid = "94ce4f54-9a6c-5748-9c1c-f9c7231a4531"
version = "1.16.1+1"

[[Libmount_jll]]
deps = ["Artifacts", "JLLWrappers", "Libdl", "Pkg"]
git-tree-sha1 = "9c30530bf0effd46e15e0fdcf2b8636e78cbbd73"
uuid = "4b2f31a3-9ecc-558c-b454-b3730dcb73e9"
version = "2.35.0+0"

[[Libtiff_jll]]
deps = ["Artifacts", "JLLWrappers", "JpegTurbo_jll", "Libdl", "Pkg", "Zlib_jll", "Zstd_jll"]
git-tree-sha1 = "340e257aada13f95f98ee352d316c3bed37c8ab9"
uuid = "89763e89-9b03-5906-acba-b20f662cd828"
version = "4.3.0+0"

[[Libuuid_jll]]
deps = ["Artifacts", "JLLWrappers", "Libdl", "Pkg"]
git-tree-sha1 = "7f3efec06033682db852f8b3bc3c1d2b0a0ab066"
uuid = "38a345b3-de98-5d2b-a5d3-14cd9215e700"
version = "2.36.0+0"

[[LinearAlgebra]]
deps = ["Libdl"]
uuid = "37e2e46d-f89d-539d-b4ee-838fcccc9c8e"

[[LogExpFunctions]]
deps = ["ChainRulesCore", "DocStringExtensions", "IrrationalConstants", "LinearAlgebra"]
git-tree-sha1 = "34dc30f868e368f8a17b728a1238f3fcda43931a"
uuid = "2ab3a3ac-af41-5b50-aa03-7779005ae688"
version = "0.3.3"

[[Logging]]
uuid = "56ddb016-857b-54e1-b83d-db4d58db5568"

[[MacroTools]]
deps = ["Markdown", "Random"]
git-tree-sha1 = "5a5bc6bf062f0f95e62d0fe0a2d99699fed82dd9"
uuid = "1914dd2f-81c6-5fcd-8719-6d5c9610ff09"
version = "0.5.8"

[[Markdown]]
deps = ["Base64"]
uuid = "d6f4376e-aef5-505a-96c1-9c027394607a"

[[MbedTLS]]
deps = ["Dates", "MbedTLS_jll", "Random", "Sockets"]
git-tree-sha1 = "1c38e51c3d08ef2278062ebceade0e46cefc96fe"
uuid = "739be429-bea8-5141-9913-cc70e7f3736d"
version = "1.0.3"

[[MbedTLS_jll]]
deps = ["Artifacts", "Libdl"]
uuid = "c8ffd9c3-330d-5841-b78e-0817d7145fa1"

[[Measurements]]
deps = ["Calculus", "LinearAlgebra", "Printf", "RecipesBase", "Requires"]
git-tree-sha1 = "31c8c0569b914111c94dd31149265ed47c238c5b"
uuid = "eff96d63-e80a-5855-80a2-b1b0885c5ab7"
version = "2.6.0"

[[Measures]]
git-tree-sha1 = "e498ddeee6f9fdb4551ce855a46f54dbd900245f"
uuid = "442fdcdd-2543-5da2-b0f3-8c86c306513e"
version = "0.3.1"

[[Missings]]
deps = ["DataAPI"]
git-tree-sha1 = "bf210ce90b6c9eed32d25dbcae1ebc565df2687f"
uuid = "e1d29d7a-bbdc-5cf2-9ac0-f12de2c33e28"
version = "1.0.2"

[[Mmap]]
uuid = "a63ad114-7e13-5084-954f-fe012c677804"

[[MozillaCACerts_jll]]
uuid = "14a3606d-f60d-562e-9121-12d972cd8159"

[[NaNMath]]
git-tree-sha1 = "bfe47e760d60b82b66b61d2d44128b62e3a369fb"
uuid = "77ba4419-2d1f-58cd-9bb1-8ffee604a2e3"
version = "0.3.5"

[[NetworkOptions]]
uuid = "ca575930-c2e3-43a9-ace4-1e988b2c1908"

[[Ogg_jll]]
deps = ["Artifacts", "JLLWrappers", "Libdl", "Pkg"]
git-tree-sha1 = "7937eda4681660b4d6aeeecc2f7e1c81c8ee4e2f"
uuid = "e7412a2a-1a6e-54c0-be00-318e2571c051"
version = "1.3.5+0"

[[OpenLibm_jll]]
deps = ["Artifacts", "Libdl"]
uuid = "05823500-19ac-5b8b-9628-191a04bc5112"

[[OpenSSL_jll]]
deps = ["Artifacts", "JLLWrappers", "Libdl", "Pkg"]
git-tree-sha1 = "15003dcb7d8db3c6c857fda14891a539a8f2705a"
uuid = "458c3c95-2e84-50aa-8efc-19380b2a3a95"
version = "1.1.10+0"

[[OpenSpecFun_jll]]
deps = ["Artifacts", "CompilerSupportLibraries_jll", "JLLWrappers", "Libdl", "Pkg"]
git-tree-sha1 = "13652491f6856acfd2db29360e1bbcd4565d04f1"
uuid = "efe28fd5-8261-553b-a9e1-b2916fc3738e"
version = "0.5.5+0"

[[Opus_jll]]
deps = ["Artifacts", "JLLWrappers", "Libdl", "Pkg"]
git-tree-sha1 = "51a08fb14ec28da2ec7a927c4337e4332c2a4720"
uuid = "91d4177d-7536-5919-b921-800302f37372"
version = "1.3.2+0"

[[OrderedCollections]]
git-tree-sha1 = "85f8e6578bf1f9ee0d11e7bb1b1456435479d47c"
uuid = "bac558e1-5e72-5ebc-8fee-abe8a469f55d"
version = "1.4.1"

[[PCRE_jll]]
deps = ["Artifacts", "JLLWrappers", "Libdl", "Pkg"]
git-tree-sha1 = "b2a7af664e098055a7529ad1a900ded962bca488"
uuid = "2f80f16e-611a-54ab-bc61-aa92de5b98fc"
version = "8.44.0+0"

[[Parameters]]
deps = ["OrderedCollections", "UnPack"]
git-tree-sha1 = "34c0e9ad262e5f7fc75b10a9952ca7692cfc5fbe"
uuid = "d96e819e-fc66-5662-9728-84c9c7592b0a"
version = "0.12.3"

[[Parsers]]
deps = ["Dates"]
git-tree-sha1 = "9d8c00ef7a8d110787ff6f170579846f776133a9"
uuid = "69de0a69-1ddd-5017-9359-2bf0b02dc9f0"
version = "2.0.4"

[[Pixman_jll]]
deps = ["Artifacts", "JLLWrappers", "Libdl", "Pkg"]
git-tree-sha1 = "b4f5d02549a10e20780a24fce72bea96b6329e29"
uuid = "30392449-352a-5448-841d-b1acce4e97dc"
version = "0.40.1+0"

[[Pkg]]
deps = ["Artifacts", "Dates", "Downloads", "LibGit2", "Libdl", "Logging", "Markdown", "Printf", "REPL", "Random", "SHA", "Serialization", "TOML", "Tar", "UUIDs", "p7zip_jll"]
uuid = "44cfe95a-1eb2-52ea-b672-e2afdf69b78f"

[[PlotThemes]]
deps = ["PlotUtils", "Requires", "Statistics"]
git-tree-sha1 = "a3a964ce9dc7898193536002a6dd892b1b5a6f1d"
uuid = "ccf2f8ad-2431-5c83-bf29-c5338b663b6a"
version = "2.0.1"

[[PlotUtils]]
deps = ["ColorSchemes", "Colors", "Dates", "Printf", "Random", "Reexport", "Statistics"]
git-tree-sha1 = "2537ed3c0ed5e03896927187f5f2ee6a4ab342db"
uuid = "995b91a9-d308-5afd-9ec6-746e21dbc043"
version = "1.0.14"

[[Plots]]
deps = ["Base64", "Contour", "Dates", "Downloads", "FFMPEG", "FixedPointNumbers", "GR", "GeometryBasics", "JSON", "Latexify", "LinearAlgebra", "Measures", "NaNMath", "PlotThemes", "PlotUtils", "Printf", "REPL", "Random", "RecipesBase", "RecipesPipeline", "Reexport", "Requires", "Scratch", "Showoff", "SparseArrays", "Statistics", "StatsBase", "UUIDs"]
git-tree-sha1 = "457b13497a3ea4deb33d273a6a5ea15c25c0ebd9"
uuid = "91a5bcdd-55d7-5caf-9e0b-520d859cae80"
version = "1.22.2"

[[PlutoUI]]
deps = ["Base64", "Dates", "HypertextLiteral", "IOCapture", "InteractiveUtils", "JSON", "Logging", "Markdown", "Random", "Reexport", "UUIDs"]
git-tree-sha1 = "0c3e067931708fa5641247affc1a1aceb53fff06"
uuid = "7f904dfe-b85e-4ff6-b463-dae2292396a8"
version = "0.7.11"

[[Preferences]]
deps = ["TOML"]
git-tree-sha1 = "00cfd92944ca9c760982747e9a1d0d5d86ab1e5a"
uuid = "21216c6a-2e73-6563-6e65-726566657250"
version = "1.2.2"

[[Printf]]
deps = ["Unicode"]
uuid = "de0858da-6303-5e67-8744-51eddeeeb8d7"

[[Profile]]
deps = ["Printf"]
uuid = "9abbd945-dff8-562f-b5e8-e1ebf5ef1b79"

[[ProgressMeter]]
deps = ["Distributed", "Printf"]
git-tree-sha1 = "afadeba63d90ff223a6a48d2009434ecee2ec9e8"
uuid = "92933f4c-e287-5a05-a399-4b506db050ca"
version = "1.7.1"

[[Qt5Base_jll]]
deps = ["Artifacts", "CompilerSupportLibraries_jll", "Fontconfig_jll", "Glib_jll", "JLLWrappers", "Libdl", "Libglvnd_jll", "OpenSSL_jll", "Pkg", "Xorg_libXext_jll", "Xorg_libxcb_jll", "Xorg_xcb_util_image_jll", "Xorg_xcb_util_keysyms_jll", "Xorg_xcb_util_renderutil_jll", "Xorg_xcb_util_wm_jll", "Zlib_jll", "xkbcommon_jll"]
git-tree-sha1 = "ad368663a5e20dbb8d6dc2fddeefe4dae0781ae8"
uuid = "ea2cea3b-5b76-57ae-a6ef-0a8af62496e1"
version = "5.15.3+0"

[[REPL]]
deps = ["InteractiveUtils", "Markdown", "Sockets", "Unicode"]
uuid = "3fa0cd96-eef1-5676-8a61-b3b8758bbffb"

[[Random]]
deps = ["Serialization"]
uuid = "9a3f8284-a2c9-5f02-9a11-845980a1fd5c"

[[RecipesBase]]
git-tree-sha1 = "44a75aa7a527910ee3d1751d1f0e4148698add9e"
uuid = "3cdcf5f2-1ef4-517c-9805-6587b60abb01"
version = "1.1.2"

[[RecipesPipeline]]
deps = ["Dates", "NaNMath", "PlotUtils", "RecipesBase"]
git-tree-sha1 = "7ad0dfa8d03b7bcf8c597f59f5292801730c55b8"
uuid = "01d81517-befc-4cb6-b9ec-a95719d0359c"
version = "0.4.1"

[[Reexport]]
git-tree-sha1 = "45e428421666073eab6f2da5c9d310d99bb12f9b"
uuid = "189a3867-3050-52da-a836-e630ba90ab69"
version = "1.2.2"

[[Requires]]
deps = ["UUIDs"]
git-tree-sha1 = "4036a3bd08ac7e968e27c203d45f5fff15020621"
uuid = "ae029012-a4dd-5104-9daa-d747884805df"
version = "1.1.3"

[[SHA]]
uuid = "ea8e919c-243c-51af-8825-aaa63cd721ce"

[[Scratch]]
deps = ["Dates"]
git-tree-sha1 = "0b4b7f1393cff97c33891da2a0bf69c6ed241fda"
uuid = "6c6a2e73-6563-6170-7368-637461726353"
version = "1.1.0"

[[Serialization]]
uuid = "9e88b42a-f829-5b0c-bbe9-9e923198166b"

[[Setfield]]
deps = ["ConstructionBase", "Future", "MacroTools", "Requires"]
git-tree-sha1 = "fca29e68c5062722b5b4435594c3d1ba557072a3"
uuid = "efcf1570-3423-57d1-acb7-fd33fddbac46"
version = "0.7.1"

[[SharedArrays]]
deps = ["Distributed", "Mmap", "Random", "Serialization"]
uuid = "1a1011a3-84de-559e-8e89-a11a2f7dc383"

[[Showoff]]
deps = ["Dates", "Grisu"]
git-tree-sha1 = "91eddf657aca81df9ae6ceb20b959ae5653ad1de"
uuid = "992d4aef-0814-514b-bc4d-f2e9a6c4116f"
version = "1.0.3"

[[Sockets]]
uuid = "6462fe0b-24de-5631-8697-dd941f90decc"

[[SortingAlgorithms]]
deps = ["DataStructures"]
git-tree-sha1 = "b3363d7460f7d098ca0912c69b082f75625d7508"
uuid = "a2af1166-a08f-5f64-846c-94a0d3cef48c"
version = "1.0.1"

[[SparseArrays]]
deps = ["LinearAlgebra", "Random"]
uuid = "2f01184e-e22b-5df5-ae63-d93ebab69eaf"

[[SpecialFunctions]]
deps = ["ChainRulesCore", "LogExpFunctions", "OpenLibm_jll", "OpenSpecFun_jll"]
git-tree-sha1 = "ad42c30a6204c74d264692e633133dcea0e8b14e"
uuid = "276daf66-3868-5448-9aa4-cd146d93841b"
version = "1.6.2"

[[StaticArrays]]
deps = ["LinearAlgebra", "Random", "Statistics"]
git-tree-sha1 = "3240808c6d463ac46f1c1cd7638375cd22abbccb"
uuid = "90137ffa-7385-5640-81b9-e52037218182"
version = "1.2.12"

[[Statistics]]
deps = ["LinearAlgebra", "SparseArrays"]
uuid = "10745b16-79ce-11e8-11f9-7d13ad32a3b2"

[[StatsAPI]]
git-tree-sha1 = "1958272568dc176a1d881acb797beb909c785510"
uuid = "82ae8749-77ed-4fe6-ae5f-f523153014b0"
version = "1.0.0"

[[StatsBase]]
deps = ["DataAPI", "DataStructures", "LinearAlgebra", "Missings", "Printf", "Random", "SortingAlgorithms", "SparseArrays", "Statistics", "StatsAPI"]
git-tree-sha1 = "8cbbc098554648c84f79a463c9ff0fd277144b6c"
uuid = "2913bbd2-ae8a-5f71-8c99-4fb6c76f3a91"
version = "0.33.10"

[[StructArrays]]
deps = ["Adapt", "DataAPI", "StaticArrays", "Tables"]
git-tree-sha1 = "2ce41e0d042c60ecd131e9fb7154a3bfadbf50d3"
uuid = "09ab397b-f2b6-538f-b94a-2f83cf4a842a"
version = "0.6.3"

[[TOML]]
deps = ["Dates"]
uuid = "fa267f1f-6049-4f14-aa54-33bafae1ed76"

[[TableTraits]]
deps = ["IteratorInterfaceExtensions"]
git-tree-sha1 = "c06b2f539df1c6efa794486abfb6ed2022561a39"
uuid = "3783bdb8-4a98-5b6b-af9a-565f29a5fe9c"
version = "1.0.1"

[[Tables]]
deps = ["DataAPI", "DataValueInterfaces", "IteratorInterfaceExtensions", "LinearAlgebra", "TableTraits", "Test"]
git-tree-sha1 = "1162ce4a6c4b7e31e0e6b14486a6986951c73be9"
uuid = "bd369af6-aec1-5ad0-b16a-f7cc5008161c"
version = "1.5.2"

[[Tar]]
deps = ["ArgTools", "SHA"]
uuid = "a4e569a6-e804-4fa4-b0f3-eef7a1d5b13e"

[[Test]]
deps = ["InteractiveUtils", "Logging", "Random", "Serialization"]
uuid = "8dfed614-e22c-5e08-85e1-65c5234f0b40"

[[URIs]]
git-tree-sha1 = "97bbe755a53fe859669cd907f2d96aee8d2c1355"
uuid = "5c2747f8-b7ea-4ff2-ba2e-563bfd36b1d4"
version = "1.3.0"

[[UUIDs]]
deps = ["Random", "SHA"]
uuid = "cf7118a7-6976-5b1a-9a39-7adc72f591a4"

[[UnPack]]
git-tree-sha1 = "387c1f73762231e86e0c9c5443ce3b4a0a9a0c2b"
uuid = "3a884ed6-31ef-47d7-9d2a-63182c4928ed"
version = "1.0.2"

[[Unicode]]
uuid = "4ec0a83e-493e-50e2-b9ac-8f72acf5a8f5"

[[Wayland_jll]]
deps = ["Artifacts", "Expat_jll", "JLLWrappers", "Libdl", "Libffi_jll", "Pkg", "XML2_jll"]
git-tree-sha1 = "3e61f0b86f90dacb0bc0e73a0c5a83f6a8636e23"
uuid = "a2964d1f-97da-50d4-b82a-358c7fce9d89"
version = "1.19.0+0"

[[Wayland_protocols_jll]]
deps = ["Artifacts", "JLLWrappers", "Libdl", "Pkg", "Wayland_jll"]
git-tree-sha1 = "2839f1c1296940218e35df0bbb220f2a79686670"
uuid = "2381bf8a-dfd0-557d-9999-79630e7b1b91"
version = "1.18.0+4"

[[XML2_jll]]
deps = ["Artifacts", "JLLWrappers", "Libdl", "Libiconv_jll", "Pkg", "Zlib_jll"]
git-tree-sha1 = "1acf5bdf07aa0907e0a37d3718bb88d4b687b74a"
uuid = "02c8fc9c-b97f-50b9-bbe4-9be30ff0a78a"
version = "2.9.12+0"

[[XSLT_jll]]
deps = ["Artifacts", "JLLWrappers", "Libdl", "Libgcrypt_jll", "Libgpg_error_jll", "Libiconv_jll", "Pkg", "XML2_jll", "Zlib_jll"]
git-tree-sha1 = "91844873c4085240b95e795f692c4cec4d805f8a"
uuid = "aed1982a-8fda-507f-9586-7b0439959a61"
version = "1.1.34+0"

[[Xorg_libX11_jll]]
deps = ["Artifacts", "JLLWrappers", "Libdl", "Pkg", "Xorg_libxcb_jll", "Xorg_xtrans_jll"]
git-tree-sha1 = "5be649d550f3f4b95308bf0183b82e2582876527"
uuid = "4f6342f7-b3d2-589e-9d20-edeb45f2b2bc"
version = "1.6.9+4"

[[Xorg_libXau_jll]]
deps = ["Artifacts", "JLLWrappers", "Libdl", "Pkg"]
git-tree-sha1 = "4e490d5c960c314f33885790ed410ff3a94ce67e"
uuid = "0c0b7dd1-d40b-584c-a123-a41640f87eec"
version = "1.0.9+4"

[[Xorg_libXcursor_jll]]
deps = ["Artifacts", "JLLWrappers", "Libdl", "Pkg", "Xorg_libXfixes_jll", "Xorg_libXrender_jll"]
git-tree-sha1 = "12e0eb3bc634fa2080c1c37fccf56f7c22989afd"
uuid = "935fb764-8cf2-53bf-bb30-45bb1f8bf724"
version = "1.2.0+4"

[[Xorg_libXdmcp_jll]]
deps = ["Artifacts", "JLLWrappers", "Libdl", "Pkg"]
git-tree-sha1 = "4fe47bd2247248125c428978740e18a681372dd4"
uuid = "a3789734-cfe1-5b06-b2d0-1dd0d9d62d05"
version = "1.1.3+4"

[[Xorg_libXext_jll]]
deps = ["Artifacts", "JLLWrappers", "Libdl", "Pkg", "Xorg_libX11_jll"]
git-tree-sha1 = "b7c0aa8c376b31e4852b360222848637f481f8c3"
uuid = "1082639a-0dae-5f34-9b06-72781eeb8cb3"
version = "1.3.4+4"

[[Xorg_libXfixes_jll]]
deps = ["Artifacts", "JLLWrappers", "Libdl", "Pkg", "Xorg_libX11_jll"]
git-tree-sha1 = "0e0dc7431e7a0587559f9294aeec269471c991a4"
uuid = "d091e8ba-531a-589c-9de9-94069b037ed8"
version = "5.0.3+4"

[[Xorg_libXi_jll]]
deps = ["Artifacts", "JLLWrappers", "Libdl", "Pkg", "Xorg_libXext_jll", "Xorg_libXfixes_jll"]
git-tree-sha1 = "89b52bc2160aadc84d707093930ef0bffa641246"
uuid = "a51aa0fd-4e3c-5386-b890-e753decda492"
version = "1.7.10+4"

[[Xorg_libXinerama_jll]]
deps = ["Artifacts", "JLLWrappers", "Libdl", "Pkg", "Xorg_libXext_jll"]
git-tree-sha1 = "26be8b1c342929259317d8b9f7b53bf2bb73b123"
uuid = "d1454406-59df-5ea1-beac-c340f2130bc3"
version = "1.1.4+4"

[[Xorg_libXrandr_jll]]
deps = ["Artifacts", "JLLWrappers", "Libdl", "Pkg", "Xorg_libXext_jll", "Xorg_libXrender_jll"]
git-tree-sha1 = "34cea83cb726fb58f325887bf0612c6b3fb17631"
uuid = "ec84b674-ba8e-5d96-8ba1-2a689ba10484"
version = "1.5.2+4"

[[Xorg_libXrender_jll]]
deps = ["Artifacts", "JLLWrappers", "Libdl", "Pkg", "Xorg_libX11_jll"]
git-tree-sha1 = "19560f30fd49f4d4efbe7002a1037f8c43d43b96"
uuid = "ea2f1a96-1ddc-540d-b46f-429655e07cfa"
version = "0.9.10+4"

[[Xorg_libpthread_stubs_jll]]
deps = ["Artifacts", "JLLWrappers", "Libdl", "Pkg"]
git-tree-sha1 = "6783737e45d3c59a4a4c4091f5f88cdcf0908cbb"
uuid = "14d82f49-176c-5ed1-bb49-ad3f5cbd8c74"
version = "0.1.0+3"

[[Xorg_libxcb_jll]]
deps = ["Artifacts", "JLLWrappers", "Libdl", "Pkg", "XSLT_jll", "Xorg_libXau_jll", "Xorg_libXdmcp_jll", "Xorg_libpthread_stubs_jll"]
git-tree-sha1 = "daf17f441228e7a3833846cd048892861cff16d6"
uuid = "c7cfdc94-dc32-55de-ac96-5a1b8d977c5b"
version = "1.13.0+3"

[[Xorg_libxkbfile_jll]]
deps = ["Artifacts", "JLLWrappers", "Libdl", "Pkg", "Xorg_libX11_jll"]
git-tree-sha1 = "926af861744212db0eb001d9e40b5d16292080b2"
uuid = "cc61e674-0454-545c-8b26-ed2c68acab7a"
version = "1.1.0+4"

[[Xorg_xcb_util_image_jll]]
deps = ["Artifacts", "JLLWrappers", "Libdl", "Pkg", "Xorg_xcb_util_jll"]
git-tree-sha1 = "0fab0a40349ba1cba2c1da699243396ff8e94b97"
uuid = "12413925-8142-5f55-bb0e-6d7ca50bb09b"
version = "0.4.0+1"

[[Xorg_xcb_util_jll]]
deps = ["Artifacts", "JLLWrappers", "Libdl", "Pkg", "Xorg_libxcb_jll"]
git-tree-sha1 = "e7fd7b2881fa2eaa72717420894d3938177862d1"
uuid = "2def613f-5ad1-5310-b15b-b15d46f528f5"
version = "0.4.0+1"

[[Xorg_xcb_util_keysyms_jll]]
deps = ["Artifacts", "JLLWrappers", "Libdl", "Pkg", "Xorg_xcb_util_jll"]
git-tree-sha1 = "d1151e2c45a544f32441a567d1690e701ec89b00"
uuid = "975044d2-76e6-5fbe-bf08-97ce7c6574c7"
version = "0.4.0+1"

[[Xorg_xcb_util_renderutil_jll]]
deps = ["Artifacts", "JLLWrappers", "Libdl", "Pkg", "Xorg_xcb_util_jll"]
git-tree-sha1 = "dfd7a8f38d4613b6a575253b3174dd991ca6183e"
uuid = "0d47668e-0667-5a69-a72c-f761630bfb7e"
version = "0.3.9+1"

[[Xorg_xcb_util_wm_jll]]
deps = ["Artifacts", "JLLWrappers", "Libdl", "Pkg", "Xorg_xcb_util_jll"]
git-tree-sha1 = "e78d10aab01a4a154142c5006ed44fd9e8e31b67"
uuid = "c22f9ab0-d5fe-5066-847c-f4bb1cd4e361"
version = "0.4.1+1"

[[Xorg_xkbcomp_jll]]
deps = ["Artifacts", "JLLWrappers", "Libdl", "Pkg", "Xorg_libxkbfile_jll"]
git-tree-sha1 = "4bcbf660f6c2e714f87e960a171b119d06ee163b"
uuid = "35661453-b289-5fab-8a00-3d9160c6a3a4"
version = "1.4.2+4"

[[Xorg_xkeyboard_config_jll]]
deps = ["Artifacts", "JLLWrappers", "Libdl", "Pkg", "Xorg_xkbcomp_jll"]
git-tree-sha1 = "5c8424f8a67c3f2209646d4425f3d415fee5931d"
uuid = "33bec58e-1273-512f-9401-5d533626f822"
version = "2.27.0+4"

[[Xorg_xtrans_jll]]
deps = ["Artifacts", "JLLWrappers", "Libdl", "Pkg"]
git-tree-sha1 = "79c31e7844f6ecf779705fbc12146eb190b7d845"
uuid = "c5fb5394-a638-5e4d-96e5-b29de1b5cf10"
version = "1.4.0+3"

[[Zlib_jll]]
deps = ["Libdl"]
uuid = "83775a58-1f1d-513f-b197-d71354ab007a"

[[Zstd_jll]]
deps = ["Artifacts", "JLLWrappers", "Libdl", "Pkg"]
git-tree-sha1 = "cc4bf3fdde8b7e3e9fa0351bdeedba1cf3b7f6e6"
uuid = "3161d3a3-bdf6-5164-811a-617609db77b4"
version = "1.5.0+0"

[[libass_jll]]
deps = ["Artifacts", "Bzip2_jll", "FreeType2_jll", "FriBidi_jll", "HarfBuzz_jll", "JLLWrappers", "Libdl", "Pkg", "Zlib_jll"]
git-tree-sha1 = "5982a94fcba20f02f42ace44b9894ee2b140fe47"
uuid = "0ac62f75-1d6f-5e53-bd7c-93b484bb37c0"
version = "0.15.1+0"

[[libfdk_aac_jll]]
deps = ["Artifacts", "JLLWrappers", "Libdl", "Pkg"]
git-tree-sha1 = "daacc84a041563f965be61859a36e17c4e4fcd55"
uuid = "f638f0a6-7fb0-5443-88ba-1cc74229b280"
version = "2.0.2+0"

[[libpng_jll]]
deps = ["Artifacts", "JLLWrappers", "Libdl", "Pkg", "Zlib_jll"]
git-tree-sha1 = "94d180a6d2b5e55e447e2d27a29ed04fe79eb30c"
uuid = "b53b4c65-9356-5827-b1ea-8c7a1a84506f"
version = "1.6.38+0"

[[libvorbis_jll]]
deps = ["Artifacts", "JLLWrappers", "Libdl", "Ogg_jll", "Pkg"]
git-tree-sha1 = "c45f4e40e7aafe9d086379e5578947ec8b95a8fb"
uuid = "f27f6e37-5d2b-51aa-960f-b287f2bc3b7a"
version = "1.3.7+0"

[[nghttp2_jll]]
deps = ["Artifacts", "Libdl"]
uuid = "8e850ede-7688-5339-a07c-302acd2aaf8d"

[[p7zip_jll]]
deps = ["Artifacts", "Libdl"]
uuid = "3f19e933-33d8-53b3-aaab-bd5110c3b7a0"

[[x264_jll]]
deps = ["Artifacts", "JLLWrappers", "Libdl", "Pkg"]
git-tree-sha1 = "4fea590b89e6ec504593146bf8b988b2c00922b2"
uuid = "1270edf5-f2f9-52d2-97e9-ab00b5d0237a"
version = "2021.5.5+0"

[[x265_jll]]
deps = ["Artifacts", "JLLWrappers", "Libdl", "Pkg"]
git-tree-sha1 = "ee567a171cce03570d77ad3a43e90218e38937a9"
uuid = "dfaa095f-4041-5dcd-9319-2fabd8486b76"
version = "3.5.0+0"

[[xkbcommon_jll]]
deps = ["Artifacts", "JLLWrappers", "Libdl", "Pkg", "Wayland_jll", "Wayland_protocols_jll", "Xorg_libxcb_jll", "Xorg_xkeyboard_config_jll"]
git-tree-sha1 = "ece2350174195bb31de1a63bea3a41ae1aa593b6"
uuid = "d8fb68d0-12a3-5cfd-a85a-d49703b185fd"
version = "0.9.1+5"
"""

# ╔═╡ Cell order:
# ╟─a87fad48-73c1-4a08-a6f1-aae759b3c6fc
# ╟─172227c2-b27a-40db-91f4-9566c2f6cf52
# ╟─4b484cf6-4888-4f04-b3fd-94862822b0c0
# ╠═a82f22e4-f35b-461a-b481-1dff43722e44
# ╠═8c444ee4-8c77-413a-bbeb-9e5ae2428876
# ╟─a0de01b5-779a-48c0-8d61-12b02a5f527e
# ╠═f87e4036-8f82-41c7-90c1-daa5f677488d
# ╟─532eb3bc-5522-4348-afa5-5336ec6752c7
# ╟─dc5f7484-3dc3-47a7-ad4a-30f97fc14d11
# ╟─ab3ff21b-bf82-4d8c-abd1-c27418956ed8
# ╠═7a1db355-bba9-4322-9fb4-a6d7b7bdd60d
# ╟─cc49ef04-08d8-42bb-9170-9db64e275a51
# ╠═df33b999-4a42-4133-bf59-5a65240790cf
# ╠═0f52365d-34f4-46ed-923e-3ea31c6db0ca
# ╟─d00e56f2-9d3a-4dd3-80eb-3201eff39b96
# ╠═b5c09cd3-6063-4a36-96cd-2d128aa11b82
# ╟─374f239b-6470-40ed-b068-a8ecaace4f09
# ╟─7719b317-e85b-4583-b401-a8614d4b2373
# ╠═f58769a6-a656-42a3-8bc6-c204d4cfd897
# ╟─144119ad-ab88-4165-883a-f2fc2464a838
# ╠═356ac5a4-c94e-42cb-a085-0198b29c7e52
# ╟─43e6b146-ee35-40f1-b540-3da22b9e1b1b
# ╟─dd9e4332-2908-40ef-b461-6b571df56cf4
# ╠═d23b4a92-055e-4ed7-bd46-8a3c59312993
# ╠═e6e29d1e-9a93-49db-a358-6b66f0bc3433
# ╟─017cd6a8-712d-4ec5-b31f-7f1f507764f2
# ╟─b5206dd5-1f46-4437-929b-efd68393b12b
# ╠═eb5dc224-1491-11ec-1cae-d51c93cd292c
# ╟─594ba1d6-2dae-4f20-9546-f52fac17c2f0
# ╟─66c7d960-7e05-4613-84e8-2a40fe40dc3d
# ╟─e717a8d9-ccfb-4f89-b2a2-f244f108b48d
# ╠═3755a4f3-1842-4de2-965e-d294c06c54c7
# ╟─505ef5ab-f131-4ab3-a723-795b5eb5dc0f
# ╟─eab7b195-64d5-4587-8687-48a673ab091b
# ╟─34dc72dc-4864-47c0-b730-183f67e7aea3
# ╠═beeb3335-5c49-47de-a1d3-3eef5f9479f1
# ╟─02d9bf3b-708c-4293-b198-9043b334ff7e
# ╠═0967b90d-ac88-476d-a57a-7c38dfa82204
# ╟─0a5282ee-c88a-4bcc-aca2-477f28e9e04d
# ╠═b2a4a505-47ff-40bb-9a6d-a08d91c53217
# ╟─fcff6973-012a-40fc-a618-f6262266287a
# ╠═985b4ffb-7964-4b50-8c2f-e5f45f352500
# ╟─14867ffd-cde5-43f8-8399-01169ee29b73
# ╟─c4798182-de75-4b59-8be7-f7cf1051364d
# ╠═efc586a2-0946-4dc5-ab3a-3902a811f3ad
# ╟─22fb0386-d4fa-47b9-ac31-decf2731cbc1
# ╟─8e23a3ea-3039-4a5f-b37f-c4710153938e
# ╠═d42f842d-6c2a-40db-b0c4-e936244a9e7c
# ╠═1ad401b5-20b2-489b-b2aa-92f729b1d725
# ╟─2a3e7257-63ad-4761-beda-cec18b91f99c
# ╟─49b1f040-929a-4238-acd9-6554757b592c
# ╠═26d5c6e9-a903-4792-a0e0-dec1a2e86a01
# ╟─2aef27bd-dea6-4a93-9d0f-b9249c9dd2cd
# ╠═0546ee2d-b62d-4c7a-8172-ba87b3c1aea4
# ╟─4a498c18-406f-4437-b378-aa9fdc75b919
# ╟─b6dcb9a3-59e3-4eae-9399-fb072c704f1a
# ╠═e4657169-1bb2-4d4a-ac9d-adc80499d07d
# ╟─5d395353-5681-4780-983e-902fdb89eaf2
# ╠═e9376a4b-3d60-42eb-8681-cd2bcec13fe8
# ╠═c5cdd71f-5ded-482f-9205-c13f01a14d0b
# ╟─a0993a1f-60a6-45b5-815e-676c49a9f049
# ╠═4e7f8db4-b5cc-4a3e-9fa7-e62d8f2a36ac
# ╟─36da3e92-000c-4d4b-9abf-4cd588b3a354
# ╠═70eb2e0a-a5c8-4975-8f6c-589035bea29c
# ╟─b4646a29-3efd-4bd1-bffc-3575559de937
# ╠═d32743c0-fc80-406f-83c5-4528e439589a
# ╟─98478246-5940-4828-a8f1-9c9fa990676d
# ╠═310f247e-3fe8-4621-ae0b-b5ee38d2ee89
# ╟─8267220a-f06e-4761-b310-00f8ba44e4b1
# ╟─ce916139-221a-462e-877f-88212663c05e
# ╠═99b5c818-a825-4939-849e-1cade802f63d
# ╟─8e2903be-4975-4e14-84ed-6e712f47fe47
# ╟─418f31bb-81d5-459b-b402-4fd4e3f4ab27
# ╠═05402cbd-78c6-4234-8680-c351c8c37778
# ╟─4e97f24c-c237-4117-bc57-e4e88c8fb8d2
# ╠═b31da90d-7165-42de-b18d-90584affea03
# ╟─5f37640b-ffd9-4877-a78c-a699b2671919
# ╠═d87c22d1-d595-4d43-ab1c-f28d282a3485
# ╟─1d6eedfd-d013-4557-9cf2-103f8fb7b72a
# ╟─bf0a5303-f5ce-4711-b9ee-a12ce2d8a397
# ╟─c003a61d-a434-4d7b-9214-5b52aa044248
# ╠═e24ce081-e367-4feb-8a79-66b8654a0b3a
# ╟─63eb391f-0238-434a-bc3a-2fa8ed41448e
# ╠═7b9bb0fd-34a5-42e1-bc35-7259447b73d0
# ╟─6a4e0e2e-75c5-4cab-987d-3d6b62f9bb06
# ╠═c91862dd-498a-4712-8e3d-b77e088cd470
# ╠═a08d6e6d-ddc4-40aa-b7c4-93ea03191415
# ╟─a356e2cc-1cb1-457a-986c-998cf1efe008
# ╠═57141f7c-9261-4dc5-98e4-b136a15f86fc
# ╟─055e32d7-073c-40db-a267-750636b9f786
# ╠═aaa97ce4-a5ff-4332-89a2-843cee2e5b6d
# ╟─1067527e-76b7-4331-b3ab-efd72fb99dfc
# ╟─93697e4d-369b-48e9-8b28-a0ff58604d02
# ╟─c4344e64-aa22-4328-a97a-71e44bcd289f
# ╟─827bda6f-87d4-4d36-8d89-f144f4595240
# ╟─1ff4077a-4742-4c5e-a8d6-c4699469a683
# ╠═4a75498d-8f4e-406f-8b01-f6a5f153919f
# ╟─3ae783ce-d06e-4cc2-b8a3-94512e8f1490
# ╠═13e7da81-8581-4f32-9fdb-2599dd36a12c
# ╟─4870b1f3-3134-4ddc-a59d-fa806b456a23
# ╠═fda6171c-9675-4f2e-b226-7ccf100529cd
# ╟─a862f8a3-0131-4644-bc90-246bf3120790
# ╠═d6564250-f646-40de-9463-a956af1a5b1d
# ╟─eee3ac4b-4ddb-4699-b6e6-f0ffcc562c07
# ╠═107aec28-ecb5-4007-95e5-25d0a7f0c465
# ╟─1394e4c6-c371-47c0-8ca8-f0830d63d8ec
# ╠═535716e6-9c1c-4324-a4cd-b1214df3c01d
# ╟─b8edfb4e-6780-4ce7-94c1-4073ff7fa832
# ╠═b8320f78-323c-49a9-a9f9-2748d19ecb35
# ╟─92737d73-676c-4c96-a321-831ecaf37690
# ╠═931a9c5f-8f91-4e88-956b-50c0efc9c58b
# ╟─b5b96082-efde-464f-bcd4-f2e0a84befcd
# ╠═7658a32c-d3da-4ec9-9d96-0d30bb18f08c
# ╟─e61981d5-5448-45e9-81dc-320ac87ba813
# ╠═31e1bb51-c531-4c4a-8634-5caafb7e9e51
# ╠═b0b81da4-6788-45c4-b618-188a02b5e09c
# ╟─47c205c3-ceae-4e12-9ade-753df1608deb
# ╟─4cef9cea-1e84-42b9-bff6-b9a8b3bfe8da
# ╟─826693ff-9a9b-46b1-aeb3-767a5e6f9441
# ╟─d231842d-9b7a-4711-b71b-5d54041ebc1f
# ╟─53cedd26-3742-4c23-a8b8-8a1f2bdfa135
# ╟─889f837d-2e26-4261-b276-5fd91efdda6a
# ╟─670a01e3-82f8-4c7f-8577-852081d91ed7
# ╠═fce1e3e0-cdf7-453f-b913-964c10fa85a6
# ╟─69a92ac6-833c-4605-b3d0-9400e4572886
# ╠═542a9ef5-d9ee-49bd-9d31-61e28b80b5cb
# ╟─8bada25c-b586-42b4-851d-232ccca8a456
# ╠═7600c6dc-769e-4c77-8526-281a1bcec079
# ╠═29dbc47b-3697-4fdf-8f34-890ab4d0cdae
# ╟─0ee7fc18-f41f-4179-a75e-1e1d56b2db29
# ╟─e5b557d7-0952-4409-ae4c-a0c8ce736e03
# ╟─0d0374ed-5150-40e6-b5a4-9a344b6ca47a
# ╠═d2f4d622-8e35-4be3-b421-39b28a748cab
# ╟─f7cf613e-be9d-4f62-a778-cc4375eb99df
# ╟─5be87c6f-5c31-4d14-a8cb-4e63ef39d538
# ╟─0c07edd3-c0a1-4f72-a16a-74badb7a6123
# ╠═4fc5ef4d-e072-41f7-aef9-b42730c8313c
# ╟─19c5cc9d-8304-4e36-a3ea-a1151f28f71d
# ╠═7dcadd85-2986-4e42-aa84-67128a8f666d
# ╟─0b5c6ede-bceb-499a-a9a8-3c6a75ed340a
# ╠═91b5eac1-4799-4a72-ac6a-e2b117b787d5
# ╟─0f86ab3c-29aa-472b-8194-228c736ee940
# ╠═0b8a2292-c0d6-44e4-b560-32d9d579a008
# ╟─d6585cca-78bf-41d1-aea3-01d9831d76cb
# ╠═1b7b7d48-79d2-4317-9045-5b7e7bd073e5
# ╟─3f9dad58-294c-405c-bfc4-67855bb1e825
# ╟─30d2f39e-5df2-4f38-8032-e5f8492ba335
# ╟─6d61b58f-b88f-48f4-8bdd-0bb1a8bc1c82
# ╟─76b8695e-64dc-44bc-8938-ce22c4a9e4d0
# ╟─372637ff-9305-4d45-bf6e-e6531dadbd14
# ╟─3b2c08a6-b27e-49be-a0b0-e5cb3d5546e0
# ╠═f049ab19-7ecf-4c65-bf6d-21352c1fe767
# ╟─7280e368-c68a-48a5-91fe-93c76607c144
# ╠═755fae26-6db9-45a0-a60d-d0e9c063f8aa
# ╟─9a8d8012-ba54-4d9b-8c4c-fe6358508f2a
# ╠═ffbeae5f-8aec-4473-a446-5b73bd911733
# ╟─3738e40f-9596-469d-aa58-a4b28d8a22f8
# ╠═5f1054b8-2337-43c1-a086-26233e95d42b
# ╠═bd719619-bdd4-4c3c-8d66-1df0f210c595
# ╟─f289955b-0239-4b8d-ba08-2edf0a7284c2
# ╠═878ab5f7-28c1-4832-9c58-cb36b360766f
# ╠═a0dcd888-059d-4abe-bb6b-958d2879101c
# ╟─2634feff-7442-4d8f-b8e5-c11113136980
# ╟─b1c5b7e5-cfbf-4d93-bd79-2924d957ae14
# ╠═cd1102ac-1500-4d79-be83-72ac9180c7ce
# ╠═f21604f4-e4f7-4d43-b3d9-32f429df443e
# ╠═10826a95-16f8-416d-b8c1-0ef3347c9b20
# ╟─c46a4f97-78e4-42fd-82b3-4dc6ce99abac
# ╠═7c433791-a653-4836-91e2-084355c01d90
# ╠═410b9da4-7848-4385-bffc-a3d9bd03cf19
# ╟─c08fff28-520e-40af-951c-fe3c324f67e0
# ╟─eb0f9080-2523-4633-be21-3a2281a1629e
# ╠═357c6621-b2b8-4f30-ba41-ffc1ae6f031b
# ╟─e3cc3c77-71ad-4006-8e27-fabaa1ae9cfb
# ╠═58eb5b4b-76ad-4f7a-b86b-0494a857dca1
# ╟─574047fa-6626-4cd0-8317-32118129711e
# ╟─739c9a8a-13a5-4a33-a441-f5bc6cb35e82
# ╟─4e059cb8-6dac-450d-9f46-b3e657d9c3cf
# ╟─5ff9c89a-d999-4af2-8e7e-fb99d4948c36
# ╠═0f2f55f6-060b-475e-bef7-eaa99da4d99f
# ╠═415ad590-247b-4a5d-b21e-7af4d0c17493
# ╟─53c2e16e-b7f5-4df2-96f4-08402b5f8979
# ╠═16cdbc18-e846-4d0a-b7e6-87f07c0c52d9
# ╟─79169f89-fedc-466b-8170-fff99b98e147
# ╠═224336e2-522c-44af-b9a1-307e2ffff0f9
# ╟─06526edf-911a-4ecc-a350-6d932ca56cd5
# ╟─b5e258cd-5542-4a4c-ae0f-91c2fee426db
# ╠═c48210e0-1a04-4f84-a4e2-f6b5d34a603d
# ╟─b0dc1c2b-82b7-488d-8074-1ef9f59a15e5
# ╠═0471b987-f987-4656-b961-285e32d0a5e1
# ╟─97b8b15b-75c7-4321-999d-b067ed2a04f9
# ╟─d9f254dc-ae4a-40b3-b682-f8a501e10a2d
# ╟─591e6a9c-444c-471f-a56b-4dfbc9111989
# ╟─1f265576-824a-4764-a738-685554068079
# ╠═339487cd-8ee8-4d1d-984b-b4c5ff00bae3
# ╟─9cb29b01-7f49-4145-96d8-c8fd971fe1c8
# ╟─ac52a71b-1138-4f1b-99c3-c174d9f09187
# ╟─1e099e1f-6494-419b-8517-5bded3e18aa6
# ╟─10c86547-d4f4-4c3f-8906-ac18ce93f3b6
# ╟─2871aca3-e6b4-4a2d-868a-36562e9a274c
# ╟─2a2e9155-1c77-46fd-8502-8431573f94d0
# ╠═7c792b6b-b6ee-4e30-88d5-d0b8064f2734
# ╟─b557a646-8c3c-4dc7-8788-bf98aec8c5c0
# ╠═febe8c06-b3aa-4db1-a3ea-fdc2a81bdebd
# ╟─260d5753-6cc2-4137-8a2c-8d8a47585ecf
# ╠═a9981931-4cc9-4d16-a6d2-34b4071a84d7
# ╟─b4154fb7-e0b0-4211-8490-8a8fe47cd2da
# ╠═8ac7b1bf-c958-4eb5-8376-f802b372e796
# ╟─b5008faf-fd43-45dd-a5a1-7f51e0b4ede5
# ╠═a756dd18-fac6-4527-944e-c16d8cc4bf95
# ╠═555d1f62-b95b-4377-a8e2-9e442ee7526d
# ╠═f5510c1e-9b9f-49f0-bc7e-0fd8e79a5760
# ╠═7f556f7c-cdb0-4f91-a359-2f933bbc5b68
# ╠═5d9a40b5-4050-47d2-9855-e9b62d56e8df
# ╠═0fc843d2-ac4f-4717-a298-92a476223112
# ╠═fc7f665b-00d9-431b-a97e-d2ff7253221a
# ╠═aadf6e48-0cbf-4973-86a1-173b6648d1df
# ╠═452e1ea7-98be-4910-ba6b-c0881fb251b2
# ╟─00000000-0000-0000-0000-000000000001
# ╟─00000000-0000-0000-0000-000000000002