update since FEM is gone

Author: ChrisRackauckas

REQUIRE

@@ -2,7 +2,7 @@ julia 0.6
 RecipesBase 0.1.0
 DiffEqBase 3.0.0
 RecursiveArrayTools 0.4.0
-DiffEqPDEBase
+DiffEqPDEBase 0.4.0
 DiffEqMonteCarlo
 DiffEqNoiseProcess
 Juno

src/convergence.jl

@@ -107,19 +107,6 @@ function analyticless_test_convergence(dts::AbstractArray,prob::AbstractODEProbl
   ConvergenceSimulation(solutions,dts,auxdata=auxdata)
 end
 
-#=
-function test_convergence(dts::AbstractArray,dxs::AbstractArray,prob::AbstractHeatProblem,convergence_axis;T=1,alg=:Euler)
-  if length(dts)!=length(dxs) error("Lengths of dts!=dxs. Invalid convergence simulation") end
-  solutions = [solve(parabolic_squaremesh([0 1 0 1],dxs[i],dts[i],T,:dirichlet),prob,alg=alg) for i in eachindex(dts)]
-  auxdata = Dict(
-            :dts => [sol.fem_mesh.dt for sol in solutions],
-            :dxs => [sol.fem_mesh.dx for sol in solutions],
-            :Δμs => [sol.fem_mesh.μ  for sol in solutions],
-            :Δνs => [sol.fem_mesh.ν  for sol in solutions])
-  return(ConvergenceSimulation(solutions,convergence_axis,auxdata=auxdata))
-end
-=#
-
 function test_convergence(probs,convergence_axis,alg;kwargs...)
   ConvergenceSimulation([solve(prob,alg;kwargs...) for prob in probs],convergence_axis)
 end

src/test_solution.jl

@@ -57,12 +57,6 @@ function appxtrue(sol::AbstractODESolution,sol2::TestSolution)
   build_solution(sol,timeseries_analytic,errors)
 end
 
-function appxtrue(sol::AbstractFEMSolution,sol2::AbstractFEMSolution)
-  u_analytic = sol2[end]
-  errors = Dict(:l∞=>maximum(abs.(sol[end]-u_analytic)),:l2=>norm(sol[end]-u_analytic,2))
-  FEMSolution(sol,u_analytic,errors)
-end
-
 """
 `appxtrue(sol::AbstractODESolution,sol2::AbstractODESolution)`