Merge pull request #40 from alan-turing-institute/dev

for a 0.1.5 release

Author: ablaom

.github/workflows/CompatHelper.yml

@@ -1,16 +1,19 @@
 name: CompatHelper
+
 on:
   schedule:
     - cron: '00 00 * * *'
-  workflow_dispatch:
+
 jobs:
   CompatHelper:
     runs-on: ubuntu-latest
     steps:
+      - uses: julia-actions/setup-julia@latest
+        with:
+          version: 1.3
       - name: Pkg.add("CompatHelper")
         run: julia -e 'using Pkg; Pkg.add("CompatHelper")'
-      - name: CompatHelper.main()
+      - name: CompatHelper.main
         env:
           GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
-          COMPATHELPER_PRIV: ${{ secrets.COMPATHELPER_PRIV }}  # optional
-        run: julia -e 'using CompatHelper; CompatHelper.main()'
+        run: julia -e 'using CompatHelper; CompatHelper.main(; master_branch = "dev")'

.github/workflows/CompatHelper.yml~

@@ -1,7 +1,7 @@
 name = "NearestNeighborModels"
 uuid = "636a865e-7cf4-491e-846c-de09b730eb36"
 authors = ["Anthony D. Blaom <anthony.blaom@gmail.com>", "Sebastian Vollmer <s.vollmer.4@warwick.ac.uk>", "Thibaut Lienart <thibaut.lienart@gmail.com>", "Okon Samuel <okonsamuel50@gmail.com>"]
-version = "0.1.4"
+version = "0.1.5"
 
 [deps]
 Distances = "b4f34e82-e78d-54a5-968a-f98e89d6e8f7"
@@ -17,7 +17,7 @@ Tables = "bd369af6-aec1-5ad0-b16a-f7cc5008161c"
 [compat]
 Distances = "^0.9, ^0.10"
 FillArrays = "^0.9, ^0.10, ^0.11 "
-MLJModelInterface = "^0.3.5, ^0.4"
+MLJModelInterface = "^0.3.5, ^0.4, 1.0"
 NearestNeighbors = "^0.4"
 StatsBase = "^0.33"
 Tables = "^1.2"

Project.toml

@@ -1,7 +1,7 @@
 # NearestNeighborModels
 | [Linux] | Coverage | Documentation |
 | :------------ | :------- | :------------ |
-| [![Build Status](https://github.com/alan-turing-institute/NearestNeighborModels.jl/workflows/CI/badge.svg)](https://github.com/alan-turing-institute/NearestNeighborModels.jl/actions) | [![Coverage](https://codecov.io/gh/alan-turing-institute/NearestNeighborModels.jl/branch/master/graph/badge.svg)](https://codecov.io/gh/vollmersj/NearestNeighborModels.jl) | [![Stable](https://img.shields.io/badge/docs-stable-blue.svg)](https://alan-turing-institute.github.io/NearestNeighborModels.jl/dev/) |
+| [![Build Status](https://github.com/alan-turing-institute/NearestNeighborModels.jl/workflows/CI/badge.svg)](https://github.com/alan-turing-institute/NearestNeighborModels.jl/actions) | [![Coverage](https://codecov.io/gh/alan-turing-institute/NearestNeighborModels.jl/branch/master/graph/badge.svg)](https://codecov.io/github/alan-turing-institute/NearestNeighborModels.jl?branch=master) | [![Stable](https://img.shields.io/badge/docs-stable-blue.svg)](https://alan-turing-institute.github.io/NearestNeighborModels.jl/dev/) |
 
 
 Package providing K-nearest neighbor regressors and classifiers, for use with the [MLJ](https://alan-turing-institute.github.io/MLJ.jl/dev/) machine learning framework.
@@ -11,7 +11,7 @@ Builds on Kristoffer Carlsson's [NearestNeighbors](https://github.com/Kristoffer
 Builds on contributions of Thibaut Lienart originally residing in [MLJModels.jl](https://github.com/alan-turing-institute/MLJModels.jl/blob/98618d7be53f72054de284fa1796c5292d9071bb/src/NearestNeighbors.jl#L1).
 
 Provides the following models: `KNNRegressor`, `KNNClassifier`,
-`MultitargetKNNRegressor` and `MultitargetClassifier`.
+`MultitargetKNNRegressor` and `MultitargetKNNClassifier`.
 
 Provides a library of kernels for weighting nearest neighbors, including
 all kernels surveyed in the paper [Geler et.al (2016):

README.md

@@ -51,7 +51,7 @@ end
 ```
 ```jldoctest ex1
 julia> predict(knnc_mach, rows=test_inds)
-60-element MLJBase.UnivariateFiniteArray{Multiclass{2},String,UInt32,Float64,1}:
+60-element MLJBase.UnivariateFiniteVector{Multiclass{2}, String, UInt32, Float64}:
  UnivariateFinite{ScientificTypes.Multiclass{2}}(B=>0.315, O=>0.685)
  UnivariateFinite{ScientificTypes.Multiclass{2}}(B=>1.0, O=>0.0)
  UnivariateFinite{ScientificTypes.Multiclass{2}}(B=>1.0, O=>0.0)
@@ -163,10 +163,12 @@ end
 ```
 ```jldoctest
 julia> table_predictions = predict(multi_knnr_mach, rows=test_inds)
-Tables.MatrixTable{Array{Float64,2}}: (x1 = [0.08676552079000954, 0.38872230178556694, -0.05722812562903978, -0.08454222308611803, 0.308204271265957, 0.17017865975808633, 0.09168784117321499, 0.9096454592475096, -0.35766988886608425, 0.03524924262394776], x2 = [0.28686596450880636, 0.11163952306296969, 0.7527073177116553, 0.1250386143966635, -0.08406347433782672, -0.01368379936598659, 0.16817186394006448, -0.7882848144244573, 0.1589684822637237, -0.3016603072338923])
+Tables.MatrixTable{Matrix{Float64}} with 10 rows, 2 columns, and schema:
+ :x1  Float64
+ :x2  Float64
 
 julia> MLJ.matrix(table_predictions)
-10×2 Array{Float64,2}:
+10×2 Matrix{Float64}:
   0.0867655   0.286866
   0.388722    0.11164
  -0.0572281   0.752707
@@ -182,4 +184,4 @@ julia> MLJ.matrix(table_predictions)
 DocTestSetup = nothing
 ```
 see [MLJ docs](https://alan-turing-institute.github.io/MLJ.jl/dev/) for help on additional 
-features such as hyper-parameter tuning, performance evaluation, stacking etc.
+features such as hyper-parameter tuning, performance evaluation, stacking etc.

docs/src/index.md

@@ -152,7 +152,7 @@ end
 function _predict_knnclassifier(weights, y, idxsvec)
     # `classes` and `classes_seen` has an ordering consistent with the pool of y
     classes = @inbounds MMI.classes(y[1])
-    classes_seen = filter(in(unique(y)), classes)
+    classes_seen = filter(in(y), classes)
     nclasses = length(classes)
     nc = length(classes_seen) # Number of classes seen in `y`.