Monotonic Basin Hopping is started

julia/src/Thesis.jl

@@ -9,4 +9,5 @@ module Thesis
   include("./laguerre-conway.jl")
   include("./propagator.jl")
   include("./find_closest.jl")
+  include("./monotonic_basin_hopping.jl")
 end

julia/src/find_closest.jl

@@ -1,18 +1,20 @@
 using NLsolve
 
+export nlp_solve
+
 function treat_inputs(x::AbstractVector)
   n::Int = length(x)/3
   reshape(x,(3,n))'
 end
 
-function single_shoot(start::Vector{Float64},
-                      final::Vector{Float64},
-                      craft::Sc,
-                      μ::Float64,
-                      t0::Float64,
-                      tf::Float64,
-                      x0::AbstractVector,
-                      tol=1e-6)
+function nlp_solve(start::Vector{Float64},
+                   final::Vector{Float64},
+                   craft::Sc,
+                   μ::Float64,
+                   t0::Float64,
+                   tf::Float64,
+                   x0::AbstractVector;
+                   tol=1e-6)
 
   n::Int = length(x0)/3
   function f!(F,x)
@@ -20,6 +22,6 @@ function single_shoot(start::Vector{Float64},
     F[7:3n] .= 0.
   end
 
-  return nlsolve(f!, x0, ftol=tol, autodiff=:forward, iterations=10_000)
+  return nlsolve(f!, x0, ftol=tol, autodiff=:forward, iterations=1_000)
 
 end

julia/src/monotonic_basin_hopping.jl

@@ -0,0 +1,56 @@
+function perturb(x::AbstractVector, n::Int)
+  perturb_vector = 0.02 * rand(Float64, (3n)) .- 0.01
+  return x + perturb_vector
+end
+
+function mass_better(x_star::AbstractVector,
+                     x_current::AbstractVector,
+                     start::AbstractVector,
+                     final::AbstractVector,
+                     craft::Sc,
+                     μ::AbstractFloat,
+                     t0::AbstractFloat,
+                     tf::AbstractFloat)
+  mass_star = prop(treat_inputs(x_star), start, craft, μ, tf-t0)[2]
+  mass_current = prop(treat_inputs(x_current), start, craft, μ, tf-t0)[2]
+  return mass_star > mass_current
+end
+
+function mbh(start::AbstractVector,
+             final::AbstractVector,
+             craft::Sc,
+             μ::AbstractFloat,
+             t0::AbstractFloat,
+             tf::AbstractFloat,
+             n::Int,
+             num_iters::Int=10,
+             tol=1e-6)
+  i::Int = 0
+  archive = []
+  x_star = nlp_solve(start, final, craft, μ, t0, tf, rand(Float64,(3n)), tol=tol)
+  while converged(x_star) == false
+    x_star = nlp_solve(start, final, craft, μ, t0, tf, rand(Float64,(3n)), tol=tol)
+  end
+
+  x_current = x_star
+  push!(archive, x_current)
+
+  while i < num_iters
+    x_star = nlp_solve(start, final, craft, μ, t0, tf, perturb(x_current.zero,n), tol=tol)
+    if converged(x_star) && mass_better(x_star.zero, x_current.zero, start, final, craft, μ, t0, tf)
+      x_current = x_star
+      push!(archive, x_star)
+    else
+      while converged(x_star) == false
+        x_star = nlp_solve(start, final, craft, μ, t0, tf, rand(Float64,(3n)), tol=tol)
+      end
+      if mass_better(x_star.zero, x_current.zero, start, final, craft, μ, t0, tf)
+        x_current = x_star
+        push!(archive, x_star)
+      end
+    end
+    i += 1
+  end
+
+  return x_current, archive
+end