Added Project and Report

project/utilities/dataset.jl

@@ -0,0 +1,42 @@
+using CSV
+using DataFrames
+using LinearAlgebra
+
+@doc raw"""
+    get_dataset(file_path::String, lambda::Float64)
+
+Reads the ML-CUP-23/24 dataset from the given file path and returns the
+necessary matrices and vectors to solve the problem. The LS problem is
+solved with the standard solver in Julia.
+
+### Input
+
+- 'file_path' -- the path to the dataset file.
+- 'lambda' -- the regularization parameter.
+
+### Output
+
+A tuple containing
+
+- 'X_hat' -- the augmented matrix X.
+- 'y_hat' -- the augmented vector y.
+- 'start' -- the initial starting point.
+- 'w_star' -- the optimal solution.
+
+"""
+function get_dataset(filePath::String, λ::Real)
+    X = CSV.File(filePath; header=false) |> DataFrame
+    X = Matrix(X)
+
+    m, n = size(X)
+    X_hat = [transpose(X); λ .* I(m)]
+    y_hat = [(rand(n).*2 .-1); zeros(m)]
+
+    # initial starting point 
+    start = ones(m)
+
+    # w_star is the optimal solution provided by the julia solver 
+    w_star = X_hat \ y_hat
+
+    return X_hat, y_hat, start, w_star
+end

project/utilities/genFunc.jl

@@ -0,0 +1,91 @@
+include("dataset.jl")
+using Random
+
+datasetpath = joinpath(@__DIR__, "../data_for_testing/dataset.csv")
+VALID = [:dataset, :randDataset, :exactRandDataset, :rand, :exactRand]
+
+@doc raw"""
+genFunc(t::Symbol, [lambda::Real=1, m::Integer=1000, n::Integer=10, rng=default_rng()])
+
+Generates different matrices for the least squares problem
+
+### Input
+
+- `t` -- The type of problem to generate. Possible values are:
+`:dataset`,
+`:randDataset`,
+`:exactRandDataset`,
+`:rand`,
+`:exactRand`
+
+### Output
+
+A named tuple with the matrix ``\hat{X}``, the vector ``\hat{y}``, the starting vector and the solution vector ``w^*``.
+
+See also [`gen_dataset`](@ref).
+"""
+function genFunc(
+        t::Symbol;
+        λ::Real=1,
+        m::Integer=1000,
+        n::Integer=10,
+        rng=Random.default_rng())::NamedTuple
+    if t ∉ VALID
+        throw(ArgumentError("The type of matrix to produce is not recognized, available types: " * join(String.(VALID), ", ")))
+    end
+
+    if t == :dataset
+        tmp = get_dataset(datasetpath, λ)
+        return (; :X_hat => tmp[1], :y_hat => tmp[2], :start => tmp[3], :w_star => tmp[4])
+    end
+
+    if t == :randDataset
+        X_hat = [(rand(rng, n, m).*2 .-1); λ .* I(m)]
+        y_hat = [(rand(rng, n).*2 .-1); zeros(m)]
+
+        # initial starting point 
+        start = ones(m)
+
+        # w_star is the optimal solution of the julia solver 
+        w_star = X_hat \ y_hat
+
+        return (; :X_hat => X_hat, :y_hat => y_hat, :start => start, :w_star => w_star)
+    end
+
+    if t == :exactRandDataset
+        X_hat = [(rand(rng, n, m).*2 .-1); λ .* I(m)]
+        w_star = rand(rng, m).*2 .-1
+
+        # initial starting point 
+        start = ones(m)
+
+        y_hat = X_hat * w_star
+
+        return (; :X_hat => X_hat, :y_hat => y_hat, :start => start, :w_star => w_star)
+    end
+
+    if t == :rand
+        X_hat = rand(rng, m+n, m).*2 .-1
+        y_hat = rand(rng, m+n)
+
+        # initial starting point 
+        start = ones(m)
+
+        # w_star is the optimal solution of the julia solver 
+        w_star = X_hat \ y_hat
+
+        return (; :X_hat => X_hat, :y_hat => y_hat, :start => start, :w_star => w_star)
+    end
+
+    if t == :exactRand
+        X_hat = rand(rng, m+n, m).*2 .-1
+        w_star = rand(rng, m).*2 .-1
+
+        # initial starting point 
+        start = ones(m)
+
+        y_hat = X_hat * w_star
+
+        return (; :X_hat => X_hat, :y_hat => y_hat, :start => start, :w_star => w_star)
+    end
+end