replace Array{Float64, x} with Vector and Matrix

src/asm2000.jl

@@ -57,7 +57,7 @@ function asm2000(setting::RegressionSetting)::Dict
 end
 
 
-function asm2000(X::Array{Float64,2}, y::Array{Float64,1})::Dict
+function asm2000(X::Matrix{Float64}, y::Vector{Float64})::Dict
     n, p = size(X)
     h = floor((n + p - 1) / 2)
     ltsreg = lts(X, y)

src/atkinson94.jl

@@ -59,8 +59,8 @@ function atkinson94(setting::RegressionSetting; iters = nothing, crit = 3.0)::Di
 end
 
 function atkinson94(
-    X::Array{Float64,2},
-    y::Array{Float64,1};
+    X::Matrix{Float64},
+    y::Vector{Float64};
     iters = nothing,
     crit = 3.0,
 )::Dict

src/atkinsonstalactiteplot.jl

@@ -34,8 +34,8 @@ function atkinsonstalactiteplot(
 end
 
 function atkinsonstalactiteplot(
-    X::Array{Float64,2},
-    y::Array{Float64,1};
+    X::Matrix{Float64},
+    y::Vector{Float64};
     iters = nothing,
     crit = 3.0,
 )::Nothing
@@ -46,7 +46,7 @@ function atkinsonstalactiteplot(
 end
 
 function generate_stalactite_plot(
-    residuals_matrix::Array{Float64,2},
+    residuals_matrix::Matrix{Float64},
     n::Int64,
     p::Int64,
     crit,

src/bacon.jl

@@ -24,7 +24,7 @@ Two methods V1 and V2 are defined in the paper which use Mahalanobis distance or
  - `method`: The distance method to use for selecting the points for initial subset
 """
 function initial_basic_subset_multivariate_data(
-    X::Array{Float64,2},
+    X::Matrix{Float64},
     m::Int;
     method::String = "mahalanobis",
 )
@@ -53,7 +53,7 @@ It also guarantees that at least m indices are returned and that the selected in
  - `m`: The minimum number of points to include in the subset indices.
  - `distances`: The distances vector used for selecting minumum distance indices.
 """
-function select_subset(X::Array{Float64,2}, m::Int, distances::Array{Float64})
+function select_subset(X::Matrix{Float64}, m::Int, distances::Array{Float64})
     rank_x = rank(X)
     sorted_distances = sortperm(distances)
     subset = sorted_distances[1:m]
@@ -88,7 +88,7 @@ This function performs the outlier detection for multivariate data according to
  - `alpha`: The quantile used for cutoff
 """
 function bacon_multivariate_outlier_detection(
-    X::Array{Float64,2},
+    X::Matrix{Float64},
     m::Int;
     method::String = "mahalanobis",
     alpha::Float64 = 0.025,
@@ -146,7 +146,7 @@ This function computes the t distance for each point and returns the distance ve
  - `y`: The output vector
  - `subset`: The vector which denotes the points inside the subset, used to scale the residuals accordingly.
 """
-function compute_t_distance(X::Array{Float64,2}, y::Array{Float64}, subset::Array{Int64})
+function compute_t_distance(X::Matrix{Float64}, y::Array{Float64}, subset::Array{Int64})
 
     n, p = size(X)
 
@@ -187,7 +187,7 @@ This function computes the initial subset having at least m elements which are l
  - `alpha`: The quantile used for cutoff
 """
 function bacon_regression_initial_subset(
-    X::Array{Float64,2},
+    X::Matrix{Float64},
     y::Array{Float64},
     m::Int;
     method::String = "mahalanobis",
@@ -259,7 +259,7 @@ Billor, Nedret, Ali S. Hadi, and Paul F. Velleman. "BACON: blocked adaptive comp
 Computational statistics & data analysis 34.3 (2000): 279-298.
 """
 function bacon(
-    X::Array{Float64,2},
+    X::Matrix{Float64},
     y::Array{Float64};
     m::Int,
     method::String = "mahalanobis",

src/basis.jl

@@ -133,7 +133,7 @@ end
 ```julia-repl
 julia> setting = createRegressionSetting(@formula(calls ~ year), phones);
 julia> designMatrix(setting)
-24×2 Array{Float64,2}:
+24×2 Matrix{Float64}:
  1.0  50.0
  1.0  51.0
  1.0  52.0
@@ -160,10 +160,10 @@ julia> designMatrix(setting)
  1.0  73.0
 ```
 """
-function designMatrix(setting::RegressionSetting)::Array{Float64,2}
+function designMatrix(setting::RegressionSetting)::Matrix{Float64}
     mf = ModelFrame(setting.formula, setting.data)
     mm = ModelMatrix(mf)
-    return convert(Array{Float64,2}, mm.m)
+    return convert(Matrix{Float64}, mm.m)
 end
 
 
@@ -182,7 +182,7 @@ end
 ```julia-repl
 julia> setting = createRegressionSetting(@formula(calls ~ year), phones);
 julia> responseVector(setting)
-24-element Array{Float64,1}:
+24-element Vector{Float64}:
    4.4
    4.7
    4.7
@@ -209,9 +209,9 @@ julia> responseVector(setting)
   29.0
 ```
 """
-function responseVector(setting::RegressionSetting)::Array{Float64,1}
+function responseVector(setting::RegressionSetting)::Vector{Float64}
     mf = ModelFrame(setting.formula, setting.data)
-    return convert(Array{Float64,1}, setting.data[:, mf.f.lhs.sym])
+    return convert(Vector{Float64}, setting.data[:, mf.f.lhs.sym])
 end
 
 
@@ -233,7 +233,7 @@ julia> X, y = @extractRegressionSetting setting;
 julia> size(X)
  (24,2)
 julia> y
-24-element Array{Float64,1}:
+24-element Vector{Float64}:
    4.4
    4.7
    4.7
@@ -313,7 +313,7 @@ julia> find_minimum_nonzero([0.0, 0.0, 5.0, 1.0])
 1.0
 ```
 """
-function find_minimum_nonzero(arr::Array{Float64,1})
+function find_minimum_nonzero(arr::Vector{Float64})
     return minimum(filter(x -> x > 0, arr))
 end
 

src/bch.jl

@@ -75,8 +75,8 @@ end
 
 
 function bch(
-    Xdesign::Array{Float64,2},
-    y::Array{Float64,1};
+    Xdesign::Matrix{Float64},
+    y::Vector{Float64};
     alpha = 0.05,
     maxiter = 1000,
     epsilon = 0.000001,
@@ -98,7 +98,7 @@ function bch(
     c = (2.0 * (n + 2.0 * p)) / (n - 2.0 * p)
 
     estimatedvariance = 0.0
-    estimatedvariances = Array{Float64,1}(undef, 0)
+    estimatedvariances = Vector{Float64}(undef, 0)
 
     # Algorithm 2 - Step 0.a
     coordmeds = coordinatwisemedians(X)

src/bchplot.jl

@@ -42,8 +42,8 @@ function bchplot(
 end
 
 function bchplot(
-    Xdesign::Array{Float64,2},
-    y::Array{Float64,1};
+    Xdesign::Matrix{Float64},
+    y::Vector{Float64};
     alpha = 0.05,
     maxiter = 1000,
     epsilon = 0.00001,

src/ccf.jl

@@ -19,7 +19,7 @@ Perform signed gradient descent for clipped convex functions for a given regress
 
 # Arguments
 - `setting::RegressionSetting`: RegressionSetting object with a formula and dataset.
-- `starting_lambdas::Array{Float64,1}`: Starting values of weighting parameters used by signed gradient descent.
+- `starting_lambdas::Vector{Float64}`: Starting values of weighting parameters used by signed gradient descent.
 - `alpha::Float64`: Loss at which a point is labeled as an outlier (points with loss ≥ alpha will be called outliers).
 - `max_iter::Int64`: Maximum number of iterations to run signed gradient descent.
 - `beta::Float64`: Step size parameter.
@@ -81,7 +81,7 @@ Perform signed gradient descent for clipped convex functions for a given regress
 # Arguments
 - `X::Matrix{Float64}`: Design matrix of the linear model.
 - `y::Vector{Float64}`: Response vector of the linear model.
-- `starting_lambdas::Array{Float64,1}`: Starting values of weighting parameters used by signed gradient descent.
+- `starting_lambdas::Vector{Float64}`: Starting values of weighting parameters used by signed gradient descent.
 - `alpha::Float64`: Loss at which a point is labeled as an outlier. If unspecified, will be chosen as p*mean(residuals.^2), where residuals are OLS residuals.
 - `p::Float64`: Points that have squared OLS residual greater than p times the mean squared OLS residual are considered outliers.
 - `max_iter::Int64`: Maximum number of iterations to run signed gradient descent.
@@ -100,8 +100,8 @@ Barratt, S., Angeris, G. & Boyd, S. Minimizing a sum of clipped convex functions
 
 """
 function ccf(
-    X::Array{Float64,2},
-    y::Array{Float64,1};
+    X::Matrix{Float64},
+    y::Vector{Float64};
     starting_lambdas = nothing,
     alpha = nothing,
     p = 3,

src/cm97.jl

@@ -50,7 +50,7 @@ end
 
 
 
-function cm97(X::Array{Float64,2}, y::Array{Float64,1}; maxiter::Int = 1000)::Dict
+function cm97(X::Matrix{Float64}, y::Vector{Float64}; maxiter::Int = 1000)::Dict
 
     n, p = size(X)
     pbar::Float64 = p / n

src/dataimage.jl

@@ -44,7 +44,7 @@ Computational Statistics & Data Analysis 43.4 (2003): 541-552.
     
 """
 function dataimage(
-    dataMatrix::Array{Float64,2};
+    dataMatrix::Matrix{Float64};
     distance = :mahalanobis,
 )::Matrix{RGBX{Float64}}
     d = nothing

src/diagnostics.jl

@@ -141,13 +141,13 @@ end
 # Example
 ```julia-repl
 julia> mat = [1.0 2.0; 3.0 4.0; 5.0 6.0]
-3×2 Array{Float64,2}:
+3×2 Matrix{Float64}:
  1.0  2.0
  3.0  4.0
  5.0  6.0
 
 julia> coordinatwisemedians(mat)
-2-element Array{Float64,1}:
+2-element Vector{Float64}:
  3.0
  4.0
 ```
@@ -224,7 +224,7 @@ Calculate `dffit` for all observations.
 julia> reg = createRegressionSetting(@formula(calls ~ year), phones);
 
 julia> dffits(reg)
-24-element Array{Float64,1}:
+24-element Vector{Float64}:
    2.3008326745719785
    1.2189579001467337
    0.35535667547543426
@@ -305,7 +305,7 @@ Calculate Studentized residuals for a given regression setting.
 julia> reg = createRegressionSetting(@formula(calls ~ year), phones);
 
 julia> studentizedResiduals(reg)
-24-element Array{Float64,1}:
+24-element Vector{Float64}:
   0.2398783264505892
   0.1463945666608097
   0.04934549995087145
@@ -361,7 +361,7 @@ Calculate adjusted residuals for a given regression setting.
 ```julia-repl
 julia> reg = createRegressionSetting(@formula(calls ~ year), phones);
 julia> adjustedResiduals(reg)
-24-element Array{Float64,1}:
+24-element Vector{Float64}:
   13.486773572526268
    8.2307993473897
    2.774371467851612
@@ -455,7 +455,7 @@ Calculate Cook distances for all observations in a regression setting.
 julia> reg = createRegressionSetting(@formula(calls ~ year), phones);
 
 julia> cooks(reg)
-24-element Array{Float64,1}:
+24-element Vector{Float64}:
  0.005344774190779822
  0.0017088194691033689
  0.00016624914057962608
@@ -617,7 +617,7 @@ Apply DFBETA diagnostic for a given regression setting and observation index.
 ```julia-repl
 julia> setting = createRegressionSetting(@formula(calls ~ year), phones);
 julia> dfbeta(setting, 1)
-2-element Array{Float64,1}:
+2-element Vector{Float64}:
   9.643915678524024
  -0.14686166007904422
 ```

src/ga.jl

@@ -5,7 +5,7 @@ import Distributions: sample
 abstract type Chromosome end
 
 mutable struct RealChromosome <: Chromosome
-    genes::Array{Float64,1}
+    genes::Vector{Float64}
     cost::Float64
 end
 
@@ -83,8 +83,8 @@ end
 function createPopulation(
     popsize::Int,
     chsize::Int,
-    mins::Array{Float64,1},
-    maxs::Array{Float64,1},
+    mins::Vector{Float64},
+    maxs::Vector{Float64},
 )::Array{RealChromosome,1}
     pop = Array{RealChromosome,1}(undef, popsize)
     for i = 1:popsize
@@ -148,8 +148,8 @@ function ga(
     popsize::Int,
     chsize::Int,
     fcost::Function,
-    mins::Array{Float64,1},
-    maxs::Array{Float64,1},
+    mins::Vector{Float64},
+    maxs::Vector{Float64},
     pcross::Float64,
     pmutate::Float64,
     elitisim::Int,

src/gwlts.jl

@@ -50,11 +50,11 @@ function galts(setting::RegressionSetting)
     return galts(X, y)
 end
 
-function galts(X::Array{Float64,2}, y::Array{Float64,1})
+function galts(X::Matrix{Float64}, y::Vector{Float64})
     n, p = size(X)
     h = Int(floor((n + p + 1.0) / 2.0))
 
-    function fcost(genes::Array{Float64,1})
+    function fcost(genes::Vector{Float64})
         objective, _ = iterateCSteps(X, y, genes, h)
         return objective
     end

src/hadi1992.jl

@@ -28,7 +28,7 @@ Perform the sub-algorithm of handling singularity defined in Hadi (1992).
 Hadi, Ali S. "Identifying multiple outliers in multivariate data." 
 Journal of the Royal Statistical Society: Series B (Methodological) 54.3 (1992): 761-771.
  """
-function hadi1992_handle_singularity(S::Array{Float64,2})::Array{Float64,2}
+function hadi1992_handle_singularity(S::Matrix{Float64})::Matrix{Float64}
     p, _ = size(S)
     eigen_structure = eigen(S)
     values = eigen_structure.values
@@ -75,7 +75,7 @@ Dict{Any,Any} with 3 entries:
 Hadi, Ali S. "Identifying multiple outliers in multivariate data." 
 Journal of the Royal Statistical Society: Series B (Methodological) 54.3 (1992): 761-771.
 """
-function hadi1992(multivariateData::Array{Float64,2}; alpha = 0.05)
+function hadi1992(multivariateData::Matrix{Float64}; alpha = 0.05)
     n, p = size(multivariateData)
     h = Int(floor((n + p + 1.0) / 2.0))
     chi_50_quantile = quantile(Chisq(p), 0.50)

src/hadi1994.jl

@@ -48,7 +48,7 @@ Dict{Any,Any} with 3 entries:
 Hadi, Ali S. "A modification of a method for the dedection of outliers in multivariate samples"
 Journal of the Royal Statistical Society: Series B (Methodological) 56.2 (1994): 393-396.
 """
-function hadi1994(multivariateData::Array{Float64,2}; alpha = 0.05)
+function hadi1994(multivariateData::Matrix{Float64}; alpha = 0.05)
     n, p = size(multivariateData)
     h = Int(round((n + p + 1.0) / 2.0))
     cnp = (1 + ((2) / (n - 1 - 3p)) + ((p + 1) / (n - p)))^2

src/hs93.jl

@@ -44,7 +44,7 @@ function hs93initialset(setting::RegressionSetting)::Array{Int,1}
     return hs93initialset(X, y)
 end
 
-function hs93initialset(X::Array{Float64,2}, y::Array{Float64,1})::Array{Int,1}
+function hs93initialset(X::Matrix{Float64}, y::Vector{Float64})::Array{Int,1}
     _, p = size(X)
     s = p + 1
     dfs = abs.(dffits(X, y))
@@ -107,8 +107,8 @@ end
 
 
 function hs93basicsubset(
-    X::Array{Float64,2},
-    y::Array{Float64,1},
+    X::Matrix{Float64},
+    y::Vector{Float64},
     initialindices::Array{Int,1},
 )::Array{Int,1}
     n, p = size(X)
@@ -186,8 +186,8 @@ end
 
 
 function hs93(
-    X::Array{Float64,2},
-    y::Array{Float64,1};
+    X::Matrix{Float64},
+    y::Vector{Float64};
     alpha = 0.05,
     basicsubsetindices = nothing,
 )