diff --git a/Manifest.toml b/Manifest.toml
index 804c596..b172a1b 100644
--- a/Manifest.toml
+++ b/Manifest.toml
@@ -12,31 +12,37 @@ git-tree-sha1 = "e214a9b9bd1b4e1b4f15b22c0994862b66af7ff7"
 uuid = "68821587-b530-5797-8361-c406ea357684"
 version = "3.5.0+3"
 
+[[Artifacts]]
+deps = ["Pkg"]
+git-tree-sha1 = "c30985d8821e0cd73870b17b0ed0ce6dc44cb744"
+uuid = "56f22d72-fd6d-98f1-02f0-08ddc0907c33"
+version = "1.3.0"
+
 [[Base64]]
 uuid = "2a0f44e3-6c83-55bd-87e4-b1978d98bd5f"
 
 [[Compat]]
 deps = ["Base64", "Dates", "DelimitedFiles", "Distributed", "InteractiveUtils", "LibGit2", "Libdl", "LinearAlgebra", "Markdown", "Mmap", "Pkg", "Printf", "REPL", "Random", "SHA", "Serialization", "SharedArrays", "Sockets", "SparseArrays", "Statistics", "Test", "UUIDs", "Unicode"]
-git-tree-sha1 = "0198d18b28c093bef39872a22f1a897218a925f5"
+git-tree-sha1 = "919c7f3151e79ff196add81d7f4e45d91bbf420b"
 uuid = "34da2185-b29b-5c13-b0c7-acf172513d20"
-version = "3.8.0"
+version = "3.25.0"
 
 [[CompilerSupportLibraries_jll]]
-deps = ["Libdl", "Pkg"]
-git-tree-sha1 = "7c4f882c41faa72118841185afc58a2eb00ef612"
+deps = ["Artifacts", "JLLWrappers", "Libdl", "Pkg"]
+git-tree-sha1 = "8e695f735fca77e9708e795eda62afdb869cbb70"
 uuid = "e66e0078-7015-5450-92f7-15fbd957f2ae"
-version = "0.3.3+0"
+version = "0.3.4+0"
 
 [[DataAPI]]
-git-tree-sha1 = "674b67f344687a88310213ddfa8a2b3c76cc4252"
+git-tree-sha1 = "ad84f52c0b8f05aa20839484dbaf01690b41ff84"
 uuid = "9a962f9c-6df0-11e9-0e5d-c546b8b5ee8a"
-version = "1.1.0"
+version = "1.4.0"
 
 [[DataStructures]]
 deps = ["InteractiveUtils", "OrderedCollections"]
-git-tree-sha1 = "73eb18320fe3ba58790c8b8f6f89420f0a622773"
+git-tree-sha1 = "88d48e133e6d3dd68183309877eac74393daa7eb"
 uuid = "864edb3b-99cc-5e75-8d2d-829cb0a9cfe8"
-version = "0.17.11"
+version = "0.17.20"
 
 [[Dates]]
 deps = ["Printf"]
@@ -58,20 +64,25 @@ version = "0.22.6"
 
 [[FillArrays]]
 deps = ["LinearAlgebra", "Random", "SparseArrays"]
-git-tree-sha1 = "3eb5253af6186eada40de3df524a1c10f0c6bfa2"
+git-tree-sha1 = "4863cbb7910079369e258dee4add9d06ead5063a"
 uuid = "1a297f60-69ca-5386-bcde-b61e274b549b"
-version = "0.8.6"
+version = "0.8.14"
 
 [[GLMNet]]
 deps = ["Distributed", "Distributions", "Printf", "Random", "SparseArrays", "StatsBase", "glmnet_jll"]
-git-tree-sha1 = "bd82e0cc37f5061d1ce29a8376a774fba759cd68"
+git-tree-sha1 = "7b2fd9b16e415f721d72b56f8b057e6919414762"
 uuid = "8d5ece8b-de18-5317-b113-243142960cc6"
-version = "0.5.1"
+version = "0.5.2"
 
 [[InteractiveUtils]]
 deps = ["Markdown"]
 uuid = "b77e0a4c-d291-57a0-90e8-8db25a27a240"
 
+[[JLLWrappers]]
+git-tree-sha1 = "04b49c556240b62d5a799e94c63d5fc14d3c07cd"
+uuid = "692b3bcd-3c85-4b1f-b108-f13ce0eb3210"
+version = "1.1.4"
+
 [[LIBLINEAR]]
 deps = ["DelimitedFiles", "Libdl", "SparseArrays", "Test"]
 git-tree-sha1 = "42cacc29d9b4ae77b6702c181bbfa58f14d8ef7a"
@@ -104,30 +115,29 @@ uuid = "d6f4376e-aef5-505a-96c1-9c027394607a"
 
 [[Missings]]
 deps = ["DataAPI"]
-git-tree-sha1 = "de0a5ce9e5289f27df672ffabef4d1e5861247d5"
+git-tree-sha1 = "ed61674a0864832495ffe0a7e889c0da76b0f4c8"
 uuid = "e1d29d7a-bbdc-5cf2-9ac0-f12de2c33e28"
-version = "0.4.3"
+version = "0.4.4"
 
 [[Mmap]]
 uuid = "a63ad114-7e13-5084-954f-fe012c677804"
 
 [[OpenBLAS_jll]]
-deps = ["Libdl", "Pkg"]
-git-tree-sha1 = "72163ef83570d258af87ed91f301c5b51a5e08e9"
+deps = ["CompilerSupportLibraries_jll", "Libdl", "Pkg"]
+git-tree-sha1 = "0c922fd9634e358622e333fc58de61f05a048492"
 uuid = "4536629a-c528-5b80-bd46-f80d51c5b363"
-version = "0.3.9+0"
+version = "0.3.9+5"
 
 [[OpenSpecFun_jll]]
-deps = ["CompilerSupportLibraries_jll", "Libdl", "Pkg"]
-git-tree-sha1 = "d51c416559217d974a1113522d5919235ae67a87"
+deps = ["Artifacts", "CompilerSupportLibraries_jll", "JLLWrappers", "Libdl", "Pkg"]
+git-tree-sha1 = "9db77584158d0ab52307f8c04f8e7c08ca76b5b3"
 uuid = "efe28fd5-8261-553b-a9e1-b2916fc3738e"
-version = "0.5.3+3"
+version = "0.5.3+4"
 
 [[OrderedCollections]]
-deps = ["Random", "Serialization", "Test"]
-git-tree-sha1 = "c4c13474d23c60d20a67b217f1d7f22a40edf8f1"
+git-tree-sha1 = "cf59cfed2e2c12e8a2ff0a4f1e9b2cd8650da6db"
 uuid = "bac558e1-5e72-5ebc-8fee-abe8a469f55d"
-version = "1.1.0"
+version = "1.3.2"
 
 [[PDMats]]
 deps = ["Arpack", "LinearAlgebra", "SparseArrays", "SuiteSparse", "Test"]
@@ -141,9 +151,9 @@ uuid = "44cfe95a-1eb2-52ea-b672-e2afdf69b78f"
 
 [[PosDefManifold]]
 deps = ["LinearAlgebra", "Statistics"]
-git-tree-sha1 = "47e697630cb91f4e27faab613361d67f47d4a5fd"
+git-tree-sha1 = "9d2e1da251012e519228a386a39fe2aa32eee12e"
 uuid = "f45a3650-5c51-11e9-1e9a-133aa5e309cf"
-version = "0.4.7"
+version = "0.4.8"
 
 [[Printf]]
 deps = ["Unicode"]
@@ -151,9 +161,9 @@ uuid = "de0858da-6303-5e67-8744-51eddeeeb8d7"
 
 [[QuadGK]]
 deps = ["DataStructures", "LinearAlgebra"]
-git-tree-sha1 = "dc84e810393cfc6294248c9032a9cdacc14a3db4"
+git-tree-sha1 = "12fbe86da16df6679be7521dfb39fbc861e1dc7b"
 uuid = "1fd47b50-473d-5c70-9696-f719f8f3bcdc"
-version = "2.3.1"
+version = "2.4.1"
 
 [[REPL]]
 deps = ["InteractiveUtils", "Markdown", "Sockets"]
@@ -171,9 +181,9 @@ version = "0.6.1"
 
 [[Rmath_jll]]
 deps = ["Libdl", "Pkg"]
-git-tree-sha1 = "1660f8fefbf5ab9c67560513131d4e933012fc4b"
+git-tree-sha1 = "d76185aa1f421306dec73c057aa384bad74188f0"
 uuid = "f50d1b31-88e8-58de-be2c-1cc44531875f"
-version = "0.2.2+0"
+version = "0.2.2+1"
 
 [[SHA]]
 uuid = "ea8e919c-243c-51af-8825-aaa63cd721ce"
@@ -206,9 +216,9 @@ uuid = "2f01184e-e22b-5df5-ae63-d93ebab69eaf"
 
 [[SpecialFunctions]]
 deps = ["OpenSpecFun_jll"]
-git-tree-sha1 = "e19b98acb182567bcb7b75bb5d9eedf3a3b5ec6c"
+git-tree-sha1 = "d8d8b8a9f4119829410ecd706da4cc8594a1e020"
 uuid = "276daf66-3868-5448-9aa4-cd146d93841b"
-version = "0.10.0"
+version = "0.10.3"
 
 [[Statistics]]
 deps = ["LinearAlgebra", "SparseArrays"]
@@ -222,9 +232,9 @@ version = "0.32.2"
 
 [[StatsFuns]]
 deps = ["Rmath", "SpecialFunctions"]
-git-tree-sha1 = "f290ddd5fdedeadd10e961eb3f4d3340f09d030a"
+git-tree-sha1 = "3b9f665c70712af3264b61c27a7e1d62055dafd1"
 uuid = "4c63d2b9-4356-54db-8cca-17b64c39e42c"
-version = "0.9.4"
+version = "0.9.6"
 
 [[SuiteSparse]]
 deps = ["Libdl", "LinearAlgebra", "Serialization", "SparseArrays"]
diff --git a/Project.toml b/Project.toml
index 9a74a8c..3da3d5f 100644
--- a/Project.toml
+++ b/Project.toml
@@ -1,7 +1,7 @@
 name = "PosDefManifoldML"
 uuid = "a07f4532-e2c9-11e9-2ea2-6d98fe4a1f21"
 authors = ["Marco-Congedo <marco.congedo@gmail.com>"]
-version = "0.4.3"
+version = "0.4.4"
 
 [deps]
 Dates = "ade2ca70-3891-5945-98fb-dc099432e06a"
@@ -15,11 +15,11 @@ Statistics = "10745b16-79ce-11e8-11f9-7d13ad32a3b2"
 StatsBase = "2913bbd2-ae8a-5f71-8c99-4fb6c76f3a91"
 
 [compat]
-Distributions = "0.22"
+Distributions = "0.22.6"
+GLMNet = "0.5.2"
 LIBSVM = "0.4.0"
-PosDefManifold = "0.4.7"
-StatsBase = "0.32"
-GLMNet = "0.5.1"
+PosDefManifold = "0.4.8"
+StatsBase = "0.32.2"
 julia = "1"
 
 [extras]
diff --git a/docs/Manifest.toml b/docs/Manifest.toml
index ad88a12..314bc96 100644
--- a/docs/Manifest.toml
+++ b/docs/Manifest.toml
@@ -13,15 +13,21 @@ uuid = "8ba89e20-285c-5b6f-9357-94700520ee1b"
 
 [[DocStringExtensions]]
 deps = ["LibGit2", "Markdown", "Pkg", "Test"]
-git-tree-sha1 = "88bb0edb352b16608036faadcc071adda068582a"
+git-tree-sha1 = "50ddf44c53698f5e784bbebb3f4b21c5807401b1"
 uuid = "ffbed154-4ef7-542d-bbb7-c09d3a79fcae"
-version = "0.8.1"
+version = "0.8.3"
 
 [[Documenter]]
-deps = ["Base64", "Dates", "DocStringExtensions", "InteractiveUtils", "JSON", "LibGit2", "Logging", "Markdown", "REPL", "Test", "Unicode"]
-git-tree-sha1 = "646ebc3db49889ffeb4c36f89e5d82c6a26295ff"
+deps = ["Base64", "Dates", "DocStringExtensions", "IOCapture", "InteractiveUtils", "JSON", "LibGit2", "Logging", "Markdown", "REPL", "Test", "Unicode"]
+git-tree-sha1 = "a4875e0763112d6d017126f3944f4133abb342ae"
 uuid = "e30172f5-a6a5-5a46-863b-614d45cd2de4"
-version = "0.24.7"
+version = "0.25.5"
+
+[[IOCapture]]
+deps = ["Logging"]
+git-tree-sha1 = "377252859f740c217b936cebcd918a44f9b53b59"
+uuid = "b5f81e59-6552-4d32-b1f0-c071b021bf89"
+version = "0.1.1"
 
 [[InteractiveUtils]]
 deps = ["Markdown"]
@@ -29,9 +35,9 @@ uuid = "b77e0a4c-d291-57a0-90e8-8db25a27a240"
 
 [[JSON]]
 deps = ["Dates", "Mmap", "Parsers", "Unicode"]
-git-tree-sha1 = "b34d7cef7b337321e97d22242c3c2b91f476748e"
+git-tree-sha1 = "81690084b6198a2e1da36fcfda16eeca9f9f24e4"
 uuid = "682c06a0-de6a-54ab-a142-c8b1cf79cde6"
-version = "0.21.0"
+version = "0.21.1"
 
 [[LibGit2]]
 deps = ["Printf"]
@@ -51,10 +57,10 @@ uuid = "d6f4376e-aef5-505a-96c1-9c027394607a"
 uuid = "a63ad114-7e13-5084-954f-fe012c677804"
 
 [[Parsers]]
-deps = ["Dates", "Test"]
-git-tree-sha1 = "75d07cb840c300084634b4991761886d0d762724"
+deps = ["Dates"]
+git-tree-sha1 = "50c9a9ed8c714945e01cd53a21007ed3865ed714"
 uuid = "69de0a69-1ddd-5017-9359-2bf0b02dc9f0"
-version = "1.0.1"
+version = "1.0.15"
 
 [[Pkg]]
 deps = ["Dates", "LibGit2", "Libdl", "Logging", "Markdown", "Printf", "REPL", "Random", "SHA", "UUIDs"]
diff --git a/docs/Project.toml b/docs/Project.toml
index 280ce40..4287f24 100644
--- a/docs/Project.toml
+++ b/docs/Project.toml
@@ -1,8 +1,8 @@
 authors = ["Marco Congedo, Saloni Jain, Anton Andreev"]
-version = "0.4.2"
+version = "0.4.4"
 
 [deps]
 Documenter = "e30172f5-a6a5-5a46-863b-614d45cd2de4"
 
 [compat]
-Documenter = "~0.24"
+Documenter = "~0.25"
diff --git a/docs/src/MainModule.md b/docs/src/MainModule.md
index 1c9ff39..3007567 100644
--- a/docs/src/MainModule.md
+++ b/docs/src/MainModule.md
@@ -20,21 +20,21 @@ The main module does not contain functions.
 
 As typical in machine learning packages, a type is created (a `struct` in Julia) to specify a ML model. *Supertype*
 
-```
+```julia
 abstract type MLmodel end
 ```
 
 is the abstract type for all machine learning
 models. *Supertype*
 
-```
+```julia
 abstract type PDmodel<:MLmodel end
 ```
 
 is the abstract type for all machine learning
 models acting on the positive definite (PD) **manifold** (for example, see [`MDM`](@ref)). *Supertype*
 
-```
+```julia
 abstract type TSmodel<:MLmodel end
 ```
 
@@ -46,7 +46,7 @@ models acting on the **tangent space** (for example, see [`ENLR`](@ref)).
 In all concerned functions *class labels* are given as a vector of integers,
 of type
 
-```
+```julia
 IntVector=Vector{Int}
 ```
 
@@ -57,7 +57,7 @@ of classes.
 
 ### working with metrics
 
-In order to work with metrics for the manifold of positive definite matrices, make sure you install the *PosDefManifold* package. 
+In order to work with metrics for the manifold of positive definite matrices, make sure you install the *PosDefManifold* package.
 
 ### the ℍVector type
 
diff --git a/docs/src/contribute.md b/docs/src/contribute.md
index 6a729b5..052d248 100644
--- a/docs/src/contribute.md
+++ b/docs/src/contribute.md
@@ -17,13 +17,13 @@ Implementing your `ABC` model entails the following five steps:
 
 if your model acts on the manifold of PSD matrices, this will be
 
-```
+```julia
 abstract type ABCmodel<:PDmodel end
 ```
 
 if your model act on the tangent space, this will be
 
-```
+```julia
 abstract type ABCmodel<:TSmodel end
 ```
 
@@ -31,7 +31,7 @@ abstract type ABCmodel<:TSmodel end
 
 This will look like:
 
-```
+```julia
 mutable struct ABC <: ABCmodel
 		metric        :: Metric
 		defaultkwarg1 :: Type_of_defaultkwarg1
@@ -67,7 +67,7 @@ if the `ABC` model has hyperparameters.
 
 Your `fit` function declaration will look like:
 
-```
+```julia
 function fit(   model :: ABCmodel,
 			    𝐏Tr :: ℍVector,
 				# if the model acts on the tangent space use:
@@ -84,13 +84,13 @@ Here you can use as many `kwarg` arguments as you wish.
 Currently, all ML models have a `verbose` argument.
 Your `fit` function should starts with:
 
-```
+```julia
 ⌚=now() # time in milliseconds
 ℳ=deepcopy(model) # output model
 ```
 and ends with
 
-```
+```julia
 verbose && println(defaultFont, "Done in ", now()-⌚,".")
 return ℳ
 ```
@@ -110,7 +110,7 @@ which should be given as options to the user, as done for these models. See how
 
 Once you have finished, a call such as
 
-```
+```julia
 m1 = fit(ABC(), PTr, yTr)
 ```
 
@@ -123,7 +123,7 @@ to the `predict` function and return the model.
 
 Your `predict` function declaration will look like:
 
-```
+```julia
 function predict(model   :: ABCmodel,
 				𝐏Te :: Union{ℍVector, Matrix{Float64}},
 				# if the model acts on the tangent space use:
@@ -139,13 +139,13 @@ where in general here you will not need `kwarg` arguments.
 
 Your `predict` function should starts with:
 
-```
+```julia
 ⌚=now() # time in milliseconds
 ```
 
 and ends with
 
-```
+```julia
 verbose && println(defaultFont, "Done in ", now()-⌚,".")
 verbose && println(titleFont, "\nPredicted ",_what2Str(what),":", defaultFont)
 return 🃏
diff --git a/docs/src/index.md b/docs/src/index.md
index 47a58f0..69e9fcb 100644
--- a/docs/src/index.md
+++ b/docs/src/index.md
@@ -31,11 +31,11 @@ To contribute, please check the secion [How to Contribute](@ref).
 ## About the Authors
 
 [Marco Congedo](https://sites.google.com/site/marcocongedo), corresponding
-author, is a research scientist of [CNRS](http://www.cnrs.fr/en) (Centre National de la Recherche Scientifique), working at [UGA](https://www.univ-grenoble-alpes.fr/english/) (University of Grenoble Alpes). **Contact**: first name dot last name at gmail dot com
+author, is a Research Director of [CNRS](http://www.cnrs.fr/en) (Centre National de la Recherche Scientifique), working at [UGA](https://www.univ-grenoble-alpes.fr/english/) (University of Grenoble Alpes). **Contact**: first name dot last name at gmail dot com
 
 Anton Andreev is a research engineer working at the same institution.
 
-Saloni Jain is a student at the
+Saloni Jain at the time of writing the package was a student at the
 [Indian Institute of Technology, Kharagpur](http://www.iitkgp.ac.in/), India.
 
 ## Overview
diff --git a/docs/src/tutorial.md b/docs/src/tutorial.md
index 434a221..9ac77c0 100644
--- a/docs/src/tutorial.md
+++ b/docs/src/tutorial.md
@@ -14,10 +14,10 @@ Furthermore, models acting on the tangent space can take as input Euclidean feat
 
 **get data**
 
-A real data example will be added soon. For now, let us create simulated data for a **2-class example**.
+Let us create simulated data for a **2-class example**.
 First, let us create symmetric positive definite matrices (real positive definite matrices):
 
-```
+```julia
 using PosDefManifoldML, PosDefManifold
 
 PTr, PTe, yTr, yTe=gen2ClassData(10, 30, 40, 60, 80, 0.1);
@@ -40,7 +40,7 @@ The **minimum distance to mean (MDM)** classifier is an example of classifier ac
 
 An MDM model is created and fit with training data such as
 
-```
+```julia
 m = fit(MDM(Fisher), PTr, yTr)
 ```
 
@@ -50,13 +50,13 @@ as declared in the parent package [PosDefManifold](https://marco-congedo.github.
 Since the Fisher metric is the default (for all ML models),
 the above is equivalent to:
 
-```
+```julia
 m = fit(MDM(), PTr, yTr)
 ```
 
 In order to adopt another metric:
 
-```
+```julia
 m1 = fit(MDM(logEuclidean), PTr, yTr)
 ```
 
@@ -64,25 +64,25 @@ m1 = fit(MDM(logEuclidean), PTr, yTr)
 
 In order to predict the labels of unlabeled data (which we have stored in `PTe`), we invoke
 
-```
+```julia
 yPred=predict(m, PTe, :l)
 ```
 
 The prediction error in percent can be retrived with
 
-```
+```julia
 predictErr(yTe, yPred)
 ```
 
 the predicton accuracy as
 
-```
+```julia
 predictAcc(yTe, yPred)
 ```
 
 and the confusion matrix as
 
-```
+```julia
 confusionMat(yTe, yPred)
 ```
 
@@ -91,13 +91,13 @@ matrices in `PTe`.
 
 If instead we wish to estimate the probabilities for the matrices in `PTe` of belonging to each class:
 
-```
+```julia
 predict(m, PTe, :p)
 ```
 
 Finally, the output functions of the MDM are obtaine by (see [`predict`](@ref))
 
-```
+```julia
 predict(m, PTe, :f)
 ```
 
@@ -105,7 +105,7 @@ predict(m, PTe, :f)
 
 The balanced accuracy estimated by a *k-fold cross-validation* is obtained such as (10-fold by default)
 
-```
+```julia
 cv = cvAcc(MDM(), PTr, yTr)
 ```
 
@@ -113,7 +113,7 @@ Struct `cv` has been created and therein you have access to average accuracy and
 and confusion matrices for all folds. For example,
 print the average confusion matrix:
 
-```
+```julia
 cv.avgCnf
 ```
 
@@ -131,7 +131,7 @@ that the [`fit`](@ref) and [`predict`](@ref) methods for ENLR models accept opti
 
 Let us get some simulated data (see the previous example for explanations).
 
-```
+```julia
 PTr, PTe, yTr, yTe=gen2ClassData(10, 30, 40, 60, 80, 0.1);
 ```
 
@@ -141,7 +141,7 @@ PTr, PTe, yTr, yTe=gen2ClassData(10, 30, 40, 60, 80, 0.1);
 
 By default, the Fisher metric ic adopted and a lasso model is fitted. The best value for the lambda hyperparameter is found by cross-validation:
 
-```
+```julia
 m1 = fit(ENLR(), PTr, yTr; w=:balanced)
 ```
 
@@ -150,50 +150,51 @@ argument (we may as well just use `w=:b`) we have requested to compute a balance
 
 The optimal value of lambda for this training data is
 
-```
+```julia
 m1.best.lambda
 ```
 
 As in *GLMNet.jl*, the intercept and beta terms are retrived by
-```
+
+```julia
 m1.best.a0
 m1.best.betas
 ```
 
 The number of non-zero beta coefficients can be found, for example, by
 
-```
+```julia
 length(unique(m1.best.betas))-1
 ```
 
 In order to fit a ridge LR model:
 
-```
+```julia
 m2 = fit(ENLR(), PTr, yTr; w=:b, alpha=0)
 ```
 
 Values of `alpha` in range ``(0, 1)`` fit instead an elastic net LR model. In the following we also request not to standardize predictors:
 
-```
+```julia
 m3 = fit(ENLR(Fisher), PTr, yTr; w=:b, alpha=0.9, standardize=false)
 ```
 
 In order to find the regularization path we use the
 `fitType` keyword argument:
 
-```
+```julia
 m1 = fit(ENLR(Fisher), PTr, yTr; w=:b, fitType=:path)
 ```
 
 The values of lambda along the path are given by
 
-```
+```julia
 m1.path.lambda
 ```
 
 We can also find the best value of the lambda hyperparameter and the regularization path at once, calling:
 
-```
+```julia
 m1 = fit(ENLR(Fisher), PTr, yTr; w=:b, fitType=:all)
 ```
 
@@ -212,7 +213,7 @@ Note that with the last call we have done here above both the `.best` and `.path
 By default, prediction is obtained from the best model
 and we request to predict the labels:
 
-```
+```julia
 yPred=predict(m1, PTe)
 
 # prediction accuracy (in proportion)
@@ -232,13 +233,13 @@ predict(m1, PTe, :f)
 In order to request the predition of labels for all models
 in the regularization path:
 
-```
+```julia
 yPred=predict(m1, PTe, :l, :path, 0)
 ```
 
 while for a specific model in the path (e.g., the 1Oth model):
 
-```
+```julia
 yPred=predict(m1, PTe, :l, :path, 10)
 ```
 
@@ -248,14 +249,14 @@ The balanced accuracy estimated by a *k-fold cross-validation* is obtained with
 some specific optional keyword arguments for
 models acting in the tangent space, for example:
 
-```
+```julia
 cv = cvAcc(ENLR(), PTr, yTr; w=:b)
 ```
 
 In order to perform another cross-validation
 arranging the training data differently in the folds:
 
-```
+```julia
 cv = cvAcc(ENLR(), PTr, yTr; w=:b, shuffle=true)
 ```
 
@@ -283,7 +284,7 @@ besides the linear one, which has no parameter. Like for ENLR, for SVM models al
 
 Let us get some simulated data as in the previous examples.
 
-```
+```julia
 PTr, PTe, yTr, yTe=gen2ClassData(10, 30, 40, 60, 80, 0.1);
 ```
 
@@ -293,7 +294,7 @@ PTr, PTe, yTr, yTe=gen2ClassData(10, 30, 40, 60, 80, 0.1);
 
 By default, a C-Support Vector Classification model is fitted:
 
-```
+```julia
 m1 = fit(SVM(), PTr, yTr; w=:b)
 ```
 
@@ -301,19 +302,19 @@ Notice that as for the example above with for ENLR model, we have requested to c
 
 In order to fit a Nu-Support Vector Classification model:
 
-```
+```julia
 m2 = fit(SVM(), PTr, yTr; w=:b, svmType=NuSVC)
 ```
 
 For using other kernels, e.g.:
 
-```
+```julia
 m3 = fit(SVM(), PTr, yTr; w=:b, svmType=NuSVC, kernel=Linear)
 ```
 
 In the following we also request not to rescale predictors:
 
-```
+```julia
 m3 = fit(SVM(), PTr, yTr;
         w=:b, svmType=NuSVC, kernel=Linear, rescale=())
 ```
@@ -328,7 +329,7 @@ details on all available optional arguments.
 
 Just the same as for the other models:
 
-```
+```julia
 yPred=predict(m1, PTe)
 
 # prediction accuracy (in proportion)
@@ -350,6 +351,6 @@ Again, the balanced accuracy estimated by a *k-fold cross-validation* is obtaine
 some specific optional keyword arguments for
 models acting in the tangent space, for example:
 
-```
+```julia
 cv = cvAcc(SVM(), PTr, yTr; w=:b)
 ```
diff --git a/src/cv.jl b/src/cv.jl
index 352a9db..3a5344c 100644
--- a/src/cv.jl
+++ b/src/cv.jl
@@ -11,7 +11,7 @@
 #   accuracy of all machine learning models.
 
 """
-```
+```julia
 struct CVacc
     cvType    :: String
     scoring   :: Union{String, Nothing}
@@ -65,7 +65,7 @@ struct CVacc
 end
 
 """
-```
+```julia
 CVacc(s::String) =
      CVacc(s, nothing, nothing, nothing,
            nothing, nothing, nothing, nothing)
@@ -79,7 +79,7 @@ CVacc(s::String)=CVacc(s, nothing, nothing, nothing, nothing, nothing, nothing,
 
 
 """
-```
+```julia
 function cvAcc(model    :: MLmodel,
                𝐏Tr     :: ℍVector,
                yTr      :: IntVector;
@@ -151,7 +151,7 @@ for balancing weights.
 **See also**: [`fit`](@ref), [`predict`](@ref).
 
 **Examples**
-```
+```julia
 using PosDefManifoldML, PosDefManifold
 
 # generate some data
@@ -298,11 +298,12 @@ end
 
 
 """
-```
+```julia
 function cvSetup(k       :: Int,
                  nCV     :: Int;
                  shuffle :: Bool = false)
 ```
+
 Given `k` elements and a parameter `nCV`, a nCV-fold cross-validation
 is obtained defining `nCV` permutations of *k* elements
 in *nTest=k÷nCV* (integer division) elements for the test and
@@ -334,7 +335,7 @@ Return the 2-tuple with:
 - A vector of `nCV` vectors holding the indices for the corresponding test sets.
 
 **Examples**
-```
+```julia
 using PosDefManifoldML, PosDefManifold
 
 cvSetup(10, 2)
diff --git a/src/enlr.jl b/src/enlr.jl
index 16056c9..b150c93 100644
--- a/src/enlr.jl
+++ b/src/enlr.jl
@@ -11,7 +11,7 @@
 #   as specific instances.
 
 """
-```
+```julia
 abstract type ENLRmodel<:TSmodel end
 ```
 
@@ -22,7 +22,7 @@ abstract type ENLRmodel<:TSmodel end
 
 
 """
-```
+```julia
 mutable struct ENLR <: ENLRmodel
 	metric      :: Metric = Fisher;
 	alpha       :: Real = 1.0
@@ -37,6 +37,7 @@ mutable struct ENLR <: ENLRmodel
 	best        :: GLMNet.GLMNetPath
 end
 ```
+
 ENLR machine learning models are incapsulated in this
 mutable structure. Fields:
 
@@ -81,7 +82,7 @@ It holds the regularization path
 that is created when the [`fit`](@ref) function is invoked
 with optional keyword parameter `fitType` = `:path` or = `:all`:
 
-```
+```julia
 struct GLMNetPath{F<:Distribution}
     family::F                        # Binomial()
     a0::Vector{Float64}              # intercept values for each solution
@@ -101,7 +102,7 @@ cross-validation used for estimating the optimal lambda
 hyperparameter by the [`fit`](@ref) function when this is invoked
 with optional keyword parameter `fitType` = `:best` (default) or = `:all`:
 
-```
+```julia
 struct GLMNetCrossValidation
     path::GLMNetPath            # the cv path
     nfolds::Int                 # the number of folds for the cv
@@ -119,7 +120,7 @@ when the [`fit`](@ref) function is invoked.
 
 
 **Examples**:
-```
+```julia
 # Note: creating models with the default creator is possible,
 # but not useful in general.
 
@@ -181,7 +182,7 @@ end
 
 
 """
-```
+```julia
 function fit(model	:: ENLRmodel,
              𝐏Tr	 :: Union{ℍVector, Matrix{Float64}},
              yTr	:: IntVector;
@@ -415,7 +416,7 @@ for details [🎓](@ref).
 **Tutorial**: [Example using the ENLR model](@ref).
 
 **Examples**
-```
+```julia
 using PosDefManifoldML, PosDefManifold
 
 # generate some data
@@ -568,7 +569,7 @@ end
 
 
 """
-```
+```julia
 function predict(model   :: ENLRmodel,
 		𝐏Te		:: Union{ℍVector, Matrix{Float64}},
 		what		:: Symbol = :labels,
@@ -643,7 +644,7 @@ tangent space is multi-threaded.
 **See also**: [`fit`](@ref), [`cvAcc`](@ref), [`predictErr`](@ref).
 
 **Examples**
-```
+```julia
 using PosDefManifoldML, PosDefManifold
 
 # generate some data
diff --git a/src/mdm.jl b/src/mdm.jl
index eb68c4d..9b87e7f 100644
--- a/src/mdm.jl
+++ b/src/mdm.jl
@@ -19,7 +19,7 @@ abstract type MDMmodel<:PDmodel end
 
 
 """
-```
+```julia
 mutable struct MDM <: MDMmodel
     metric  :: Metric = Fisher;
     featDim :: Int
@@ -59,7 +59,7 @@ It is used to optimize the computation of distances if the
 model is fitted useing the Fisher metric (default).
 
 **Examples**:
-```
+```julia
 using PosDefManifoldML, PosDefManifold
 
 # create an empty model
@@ -92,7 +92,7 @@ end
 
 
 """
-```
+```julia
 function fit(model :: MDMmodel,
               𝐏Tr   :: ℍVector,
               yTr   :: IntVector;
@@ -149,7 +149,7 @@ without crowding the REPL.
 **See also**: [`predict`](@ref), [`cvAcc`](@ref).
 
 **Examples**
-```
+```julia
 using PosDefManifoldML, PosDefManifold
 
 # generate some data
@@ -213,7 +213,7 @@ end
 
 
 """
-```
+```julia
 function predict(model  :: MDMmodel,
                  𝐏Te    :: ℍVector,
                  what   :: Symbol = :labels;
@@ -311,7 +311,7 @@ end
 
 
 """
-```
+```julia
 function getMean(metric :: Metric,
                  𝐏      :: ℍVector;
               w        :: Vector = [],
@@ -378,7 +378,7 @@ end
 
 
 """
-```
+```julia
 function getDistances(metric :: Metric,
                       means  :: ℍVector,
                       𝐏      :: ℍVector;
diff --git a/src/svm.jl b/src/svm.jl
index 1a1f98b..26cab25 100644
--- a/src/svm.jl
+++ b/src/svm.jl
@@ -9,7 +9,7 @@
 #   and it applies SVM classification using Julia's SVM wrapper.
 
 """
-```
+```julia
 abstract type SVMmodel<:TSmodel end
 ```
 Abstract type for **Support-Vector Machine (SVM)**
@@ -18,7 +18,7 @@ learning models. See [MLmodel](@ref).
 abstract type SVMmodel<:TSmodel end
 
 """
-```
+```julia
 mutable struct SVM <: SVMmodel
 	metric		:: Metric
 	svmType		:: Type
@@ -81,7 +81,7 @@ see the documentation of the [`ENLR`](@ref) structure.
 (declared [here](https://github.com/mpastell/LIBSVM.jl/blob/master/src/LibSVMtypes.jl)).
 
 **Examples**:
-```
+```julia
 # Note: creating models with the default creator is possible,
 # but not useful in general.
 
@@ -146,7 +146,7 @@ end
 
 
 """
-```
+```julia
 function fit(model     :: SVMmodel,
                𝐏Tr     :: Union{ℍVector, Matrix{Float64}},
                yTr     :: IntVector=[];
@@ -257,7 +257,7 @@ resources on the LIBSVM package [🎓](@ref).
 **Tutorial**: [Example using SVM models](@ref).
 
 **Examples**
-```
+```julia
 using PosDefManifoldML, PosDefManifold
 
 # generate some data
@@ -377,7 +377,7 @@ end
 
 
 """
-```
+```julia
 function predict(model :: SVMmodel,
 				𝐏Te	:: Union{ℍVector, Matrix{Float64}},
 				what	:: Symbol = :labels;
@@ -403,7 +403,7 @@ will be multi-threaded.
 **See also**: [`fit`](@ref), [`cvAcc`](@ref), [`predictErr`](@ref).
 
 **Examples**
-```
+```julia
 using PosDefManifoldML, PosDefManifold
 
 # generate some data
diff --git a/src/tools.jl b/src/tools.jl
index 81c4758..d84dab2 100644
--- a/src/tools.jl
+++ b/src/tools.jl
@@ -10,7 +10,7 @@
 #   Riemannian and Euclidean machine learning classifiers.
 
 """
-```
+```julia
 function tsMap(	metric :: Metric,
 		𝐏      :: ℍVector;
 		w    	  :: Vector	= [],
@@ -84,7 +84,7 @@ are multi-threaded. Multi-threading is automatically disabled if the
 number of threads Julia is instructed to use is *<2* or *<2k*.
 
 **Examples**:
-```
+```julia
 using PosDefManifoldML
 
 # generate four random symmetric positive definite 3x3 matrices
@@ -142,7 +142,7 @@ end
 
 
 """
-```
+```julia
 function tsWeights(y::Vector{Int}; classWeights=[])
 ```
 
@@ -176,7 +176,7 @@ implicitly passing symbol `:balanced` (or just `:b`) or a tuple
 with the class weights as optional keyword argument `w`.
 
 **Examples**
-```
+```julia
 # generate some data; the classes are unbalanced
 PTr, PTe, yTr, yTe=gen2ClassData(10, 30, 40, 60, 80, 0.1)
 
@@ -194,8 +194,9 @@ m=fit(ENLR(), PTr, yTr; w=(0.5, 1.5))
 # which is equivalent to
 m=fit(ENLR(), PTr, yTr; w=tsWeights(yTr; classWeights=(0.5, 1.5)))
 
+```
 
-# This is how it works:
+This is how it works:
 
 julia> y=[1, 1, 1, 1, 2, 2]
 6-element Array{Int64,1}:
@@ -235,7 +236,6 @@ julia> tsWeights(y, classWeights=[1, 4])
 
 and, again, all weights sum up to 1
 
-```
 """
 function tsWeights(y::Vector{Int}; classWeights=[])
 
@@ -262,7 +262,7 @@ end
 
 
 """
-```
+```julia
 function gen2ClassData(n        ::  Int,
                        k1train  ::  Int,
                        k2train  ::  Int,
@@ -292,7 +292,7 @@ Return a 4-tuple with
 
 **Examples**
 
-```
+```julia
 using PosDefManifoldML
 
 PTr, PTe, yTr, yTe=gen2ClassData(10, 30, 40, 60, 80, 0.25)
@@ -355,7 +355,7 @@ end
 
 
 """
-```
+```julia
 function confusionMat(yTrue::IntVector, yPred::IntVector)
 ```
 
@@ -377,7 +377,7 @@ Therefore, the entries of the confusion matrix sum up to 1.0.
 
 **Examples**
 
-```
+```julia
 using PosDefManifoldML
 julia> confusionMat([1, 1, 1, 2, 2], [1, 1, 1, 1, 2])
 # return: [0.6 0.0; 0.2 0.2]
@@ -405,7 +405,7 @@ function confusionMat(yTrue::IntVector, yPred::IntVector)
 end
 
 """
-```
+```julia
 (1)
 function predictAcc(yTrue::IntVector, yPred::IntVector;
 		scoring:: Symbol = :b,
@@ -446,7 +446,7 @@ and their mean is taken.
 
 **Examples**
 
-```
+```julia
 using PosDefManifoldML
 julia> predictAcc([1, 1, 1, 2, 2], [1, 1, 1, 1, 2]; scoring=:a)
 # regular accuracy, return: 0.8
@@ -498,7 +498,7 @@ end
 
 
 """
-```
+```julia
 (1)
 function predictErr(yTrue::IntVector, yPred::IntVector;
 		scoring:: Symbol = :b,
@@ -530,7 +530,7 @@ predictErr(CM:: Matrix{R};
 				scoring=scoring, digits=8))≠nothing ? round(1.0-acc;
 													  digits=digits) : nothing
 """
-```
+```julia
 function rescale!(X::Matrix{T},	bounds::Tuple=(-1, 1);
 		dims::Int=1) where T<:Real
 ```