Thibaut Lienart

Add short docstrings

1

At the very least should list and explain fields. * [x] Clustering * [x] KMeans * [x] KMedoids * [ ] GaussianProcesses * [ ] GPClassifier * [x] GLM *...

This is for future references, Sklearn sometimes can process sparse data and/or return sparse data; at the moment we ignore this but in the future (once MLJ handles sparse data)...

Use of MLJBase.matrix(...; transpose=true)

1

Following comments on the KNN PR, I was wondering about what the `MLJBase.matrix(..transpose=true)` does. Unless I'm mistaken it **materializes** the transpose (i.e. the output of that call is a Matrix...

Improve how HTMLWriter determines URLs in generated HTML in fixlinks!

2

In the context of `LiveServer.servedocs()`, I sometimes get a warning such as ``` ┌ Warning: invalid local image: unresolved path in man/ls+lit.md │ link = "/assets/testlit.png" └ @ Documenter.Writers.HTMLWriter ~/.julia/packages/Documenter/yf96B/src/Writers/HTMLWriter.jl:2103...

arrow up next to the house does not seem to point to the relevant parent dir.

I don't know if you guys have had a look at [QueryVerse](https://github.com/davidanthoff/Queryverse.jl), looks like something to be considered at least for the data wrangling part. (Only it does not seem...

DecisionTree notebook

3

```julia # I would strongly suggest not running past n=4 n = 5 ``` is a bit funny. Re benchmark with Python do you 1. have some idea why DecisionTree.jl...

MLLabelUtils, MLPreproc

2

I'm under the impression that JuliaDB has its own One Hot Encoding facility so that MLlabelutils would effectively become deprecated, is this correct? Same question for MLPReprocessoing

I think the `REQUIRE` file from that package may be incomplete. `LossFunctions` (cf. error message) is a package that you should be able to add pretty easily with `Pkg.add("LossFunctions")` ?

Joris published his code on CRAN (https://cran.r-project.org/web/packages/RZigZag/) - have a look at the code - compare with PDSampler with ZZ - see if any improvements can be made