Passed all tests

Author: theogf

examples/classification_SXGPC.jl

@@ -1,5 +1,4 @@
 using Plots
-using GradDescent
 pyplot()
 #unicodeplots()
 using Distributions
@@ -49,7 +48,7 @@ Ninducingpoints = 20
  tic()
 # model = DAM.BatchXGPC(X,y;Kernels=kerns,Autotuning=true,optimizer=StandardGD(α=0.1),AutotuningFrequency=1,verbose=2,ϵ=1e-4,nEpochs=100)
 Profile.clear()
-@profile model = SparseXGPC(X,y;optimizer=Adam(α=0.5),OptimizeIndPoints=true,
+@profile model = SparseXGPC(X,y;OptimizeIndPoints=true,
 Stochastic=false,ϵ=1e-4,nEpochs=MaxIter,SmoothingWindow=10,Kernels=kerns,Autotuning=false,AutotuningFrequency=2,
 verbose=2,AdaptiveLearningRate=true,batchsize=batchsize,m=Ninducingpoints)
 ProfileView.view()

src/OMGP.jl

@@ -27,13 +27,13 @@ abstract type FullBatchModel <: NonLinearModel end
 
 export GPModel, OnlineGPModel, OfflineGPModel, SparseModel, NonLinearModel, LinearModel, FullBatchModel, GPMOdelType
 
+include("graddescent/GradDescent.jl")
 include("kernels/KernelModule.jl")
 include("kmeans/KMeansModule.jl")
 include("functions/PGSampler.jl")
 include("functions/PerturbativeCorrection.jl")
 include("functions/GPAnalysisTools.jl")
 include("functions/IO_model.jl")
-include("graddescent/GradDescent.jl")
 #Custom modules
 using .KernelModule
 using .KMeansModule

src/functions/IO_model.jl

@@ -4,7 +4,6 @@ using JLD2, FileIO
 # using OMGP
 import OMGP.KernelModule
 import OMGP.KernelModule.HyperParametersModule#: HyperParameter, OpenBound, NullBound, Interval
-import GradDescent
 export save_trained_model, save_model,load_trained_model,load_model
 #Needs to be redefined
 function save_model(filename::String,model::GPModel)

src/kernels/hyperparameters/HyperParametersModule.jl

@@ -1,5 +1,5 @@
 module HyperParametersModule
-using GradDescent
+using OMGP.GradDescent
 import Base: convert, eltype, promote_type, show, string, ==, *, /, +, -, ^, isless;
 
 export