Merge pull request #4 from lalvim/renamepkg

Rename package PLS to PLSRegressor

.travis.yml

@@ -5,7 +5,6 @@ os:
   - osx
 julia:
   - 0.6
-  - nightly
 notifications:
   email: false
 git:
@@ -27,9 +26,9 @@ addons:
 
 ## uncomment the following lines to override the default test script
 #script:
-#  - julia -e 'Pkg.clone(pwd()); Pkg.build("PLS"); Pkg.test("PLS"; coverage=true)'
+#  - julia -e 'Pkg.clone(pwd()); Pkg.build("PLSRegressor"); Pkg.test("PLSRegressor"; coverage=true)'
 after_success:
   # push coverage results to Coveralls
-  - julia -e 'cd(Pkg.dir("PLS")); Pkg.add("Coverage"); using Coverage; Coveralls.submit(Coveralls.process_folder())'
+  - julia -e 'cd(Pkg.dir("PLSRegressor")); Pkg.add("Coverage"); using Coverage; Coveralls.submit(Coveralls.process_folder())'
   # push coverage results to Codecov
-  - julia -e 'cd(Pkg.dir("PLS")); Pkg.add("Coverage"); using Coverage; Codecov.submit(Codecov.process_folder())'
+  - julia -e 'cd(Pkg.dir("PLSRegressor")); Pkg.add("Coverage"); using Coverage; Codecov.submit(Codecov.process_folder())'

LICENSE.md

@@ -1,4 +1,4 @@
-The PLS.jl package is licensed under the MIT "Expat" License:
+The PLSRegressor.jl package is licensed under the MIT "Expat" License:
 
 > Copyright (c) 2017: Leandro Alvim.
 >

README.md

@@ -1,4 +1,4 @@
-PLS.jl
+PLSRegressor.jl
 ======
 
 A Partial Least Squares Regressor package. Contains PLS1, PLS2 and Kernel PLS2 NIPALS algorithms.
@@ -9,41 +9,41 @@ Can be used mainly for regression. However, for classification task, binarizing
 |:-------------------------------:|:-----------------------------------------:|
 | [![][pkg-0.6-img]][pkg-0.6-url] | [![][travis-img]][travis-url] [![][codecov-img]][codecov-url] |
 
-[travis-img]: https://travis-ci.org/lalvim/PLS.jl.svg?branch=master
-[travis-url]: https://travis-ci.org/lalvim/PLS.jl
+[travis-img]: https://travis-ci.org/lalvim/PLSRegressor.jl.svg?branch=master
+[travis-url]: https://travis-ci.org/lalvim/PLSRegressor.jl
 
-[codecov-img]: http://codecov.io/github/lalvim/PLS.jl/coverage.svg?branch=master
-[codecov-url]: http://codecov.io/github/lalvim/PLS.jl?branch=master
+[codecov-img]: http://codecov.io/github/lalvim/PLSRegressor.jl/coverage.svg?branch=master
+[codecov-url]: http://codecov.io/github/lalvim/PLSRegressor.jl?branch=master
 
-[issues-url]: https://github.com/lalvim/PLS.jl/issues
+[issues-url]: https://github.com/lalvim/PLSRegressor.jl/issues
 
-[pkg-0.6-img]: http://pkg.julialang.org/badges/PLS_0.6.svg
-[pkg-0.6-url]: http://pkg.julialang.org/?pkg=PLS&ver=0.6
-[pkg-0.7-img]: http://pkg.julialang.org/badges/PLS_0.7.svg
-[pkg-0.7-url]: http://pkg.julialang.org/?pkg=PLS&ver=0.7
+[pkg-0.6-img]: http://pkg.julialang.org/badges/PLSRegressor_0.6.svg
+[pkg-0.6-url]: http://pkg.julialang.org/?pkg=PLSRegressor&ver=0.6
+[pkg-0.7-img]: http://pkg.julialang.org/badges/PLSRegressor_0.7.svg
+[pkg-0.7-url]: http://pkg.julialang.org/?pkg=PLSRegressor&ver=0.7
 
 Install
 =======
 
-    Pkg.add("PLS")
+    Pkg.add("PLSRegressor")
 
 Using
 =====
 
-    using PLS
+    using PLSRegressor
 
 Examples
 ========
 
-    using PLS
+    using PLSRegressor
 
     # learning a single target
     X_train        = [1 2; 2 4; 4 6.0]
     Y_train        = [4; 6; 8.0]
     X_test         = [6 8; 8 10; 10 12.0]
 
-    model          = PLS.fit(X_train,Y_train,nfactors=2)
-    Y_test         = PLS.predict(model,X_test)
+    model          = PLSRegressor.fit(X_train,Y_train,nfactors=2)
+    Y_test         = PLSRegressor.predict(model,X_test)
 
     print("[PLS1] mae error : $(mean(abs.(Y_test .- Y_pred)))")
 
@@ -53,23 +53,23 @@ Examples
     Y_train        = [2 4;4 6;6 8.0]
     X_test         = [6 8; 8 10; 10 12.0]
 
-    model          = PLS.fit(X_train,Y_train,nfactors=2)
-    Y_test         = PLS.predict(model,X_test)
+    model          = PLSRegressor.fit(X_train,Y_train,nfactors=2)
+    Y_test         = PLSRegressor.predict(model,X_test)
 
     print("[PLS2] mae error : $(mean(abs.(Y_test .- Y_pred)))")
 
     # nonlinear learning with multiple targets
-    model          = PLS.fit(X_train,Y_train,nfactors=2,kernel="rbf",width=0.1)
-    Y_test         = PLS.predict(model,X_test)
+    model          = PLSRegressor.fit(X_train,Y_train,nfactors=2,kernel="rbf",width=0.1)
+    Y_test         = PLSRegressor.predict(model,X_test)
 
     print("[KPLS] mae error : $(mean(abs.(Y_test .- Y_pred)))")
 
 
     # if you want to save your model
-    PLS.save(model,filename="/tmp/pls_model.jld")
+    PLSRegressor.save(model,filename="/tmp/pls_model.jld")
 
     # if you want to load back your model
-    model = PLS.load(filename="/tmp/pls_model.jld")
+    model = PLSRegressor.load(filename="/tmp/pls_model.jld")
 
 
 What is Implemented
@@ -87,7 +87,7 @@ What is Upcoming
 Method Description
 =======
 
-* PLS.fit - learns from input data and its related single target
+* PLSRegressor.fit - learns from input data and its related single target
     * X::Matrix{:<AbstractFloat} - A matrix that columns are the features and rows are the samples
     * Y::Vector{:<AbstractFloat} - A vector with float values.
     * nfactors::Int = 10 - The number of latent variables to explain the data.
@@ -96,8 +96,8 @@ Method Description
     * kernel::AbstractString = "rbf" - use a non linear kernel.
     * width::AbstractFloat   = 1.0 - If you want to z-score columns. Recommended if not z-scored yet.
 
-* PLS.transform - predicts using the learnt model extracted from fit.
-    * model::PLS.Model - A PLS model learnt from fit.
+* PLSRegressor.transform - predicts using the learnt model extracted from fit.
+    * model::PLSRegressor.Model - A PLS model learnt from fit.
     * X::Matrix{:<AbstractFloat} - A matrix that columns are the features and rows are the samples.
     * copydata::Bool = true - If you want to use the same input matrix or a copy.
 
@@ -123,5 +123,5 @@ regression. Chemometrics and Intelligent Laboratory Systems, 18: 251–
 License
 =======
 
-The PLS.jl is free software: you can redistribute it and/or modify it under the terms of the MIT "Expat"
+The PLSRegressor.jl is free software: you can redistribute it and/or modify it under the terms of the MIT "Expat"
 License. A copy of this license is provided in ``LICENSE.md``

appveyor.yml

@@ -41,7 +41,7 @@ build_script:
 # Need to convert from shallow to complete for Pkg.clone to work
   - IF EXIST .git\shallow (git fetch --unshallow)
   - C:\projects\julia\bin\julia -e "versioninfo();
-      Pkg.clone(pwd(), \"PLS\"); Pkg.build(\"PLS\")"
+      Pkg.clone(pwd(), \"PLSRegressor\"); Pkg.build(\"PLSRegressor\")"
 
 test_script:
-  - C:\projects\julia\bin\julia -e "Pkg.test(\"PLS\")"
+  - C:\projects\julia\bin\julia -e "Pkg.test(\"PLSRegressor\")"

experiments/curve.jl

@@ -4,7 +4,7 @@
 
 ### Example extracted from a python KPLS implementation: https://github.com/jhumphry/regressions/blob/master/examples/kpls_example.py
 
-using PLS
+using PLSRegressor
 using Gadfly
 
 srand(1)
@@ -25,8 +25,8 @@ global best_g = 10
 for g in [1,2],
     w in linspace(0.01,3,10)
     print(".")
-    model      = PLS.fit(X,Y,centralize=true,nfactors=g,kernel="rbf",width=w)
-    Y_pred     = PLS.predict(model,X)
+    model      = PLSRegressor.fit(X,Y,centralize=true,nfactors=g,kernel="rbf",width=w)
+    Y_pred     = PLSRegressor.predict(model,X)
     mae = mean(abs.(Y .- Y_pred))
     if mae < min_mae
        min_mae = mae

experiments/housing.jl

@@ -1,7 +1,7 @@
 using MultivariateStats
-using PLS
+using PLSRegressor
 
-const defdir = PLS.dir("datasets")
+const defdir = PLSRegressor.dir("datasets")
 
 function gethousingdata(dir, filename)
     url = "https://archive.ics.uci.edu/ml/machine-learning-databases/housing/housing.data"
@@ -38,8 +38,8 @@ end
 
 (xtrn, ytrn, xtst, ytst) = loaddata()
 
-model    = PLS.fit(xtrn, ytrn, nfactors = 3)
-pred     = PLS.predict(model, xtst)
+model    = PLSRegressor.fit(xtrn, ytrn, nfactors = 3)
+pred     = PLSRegressor.predict(model, xtst)
 
 
 println("[PLS] mae error :", mean(abs.(ytst .- pred)))
@@ -51,5 +51,5 @@ yp = xtst * a + b
 println("[LLS] mae error :",mean(abs.(ytst .- yp)))
 
 ### if you want to save or load model use this
-#PLS.save(model,filename="/tmp/pls_model.jld",modelname="pls_model")
-#model = PLS.load(filename="/tmp/pls_model.jld",modelname="pls_model")
+#PLSRegressor.save(model,filename="/tmp/pls_model.jld",modelname="pls_model")
+#model = PLSRegressor.load(filename="/tmp/pls_model.jld",modelname="pls_model")

src/PLS.jl → src/PLSRegressor.jl

@@ -1,5 +1,5 @@
 # Partial Least Squares (PLS1 and PLS2 NIPALS version)
-module PLS
+module PLSRegressor
 
 using JLD
 

src/method.jl

@@ -57,7 +57,7 @@ end
 
 
 """
-    transform(model::PLS.Model; X::Matrix{:<AbstractFloat}; copydata::Bool=true)
+    transform(model::PLSRegressor.Model; X::Matrix{:<AbstractFloat}; copydata::Bool=true)
 
 A Partial Least Squares predictor.
 

test/kpls_test.jl

@@ -13,8 +13,8 @@
         z_noisy  = z_pure + noise
         X        = collect(x_values)
         Y        = z_noisy #z_pure
-        model    = PLS.fit(X,Y,nfactors=1,kernel="rbf",width=0.01)
-    	Y_pred   = PLS.predict(model,X)
+        model    = PLSRegressor.fit(X,Y,nfactors=1,kernel="rbf",width=0.01)
+    	Y_pred   = PLSRegressor.predict(model,X)
         @test mean(abs.(Y .- Y_pred)) < 1e-2
 
     end
@@ -24,14 +24,14 @@
 
         X        = [1 2; 2 4; 4.0 6]
         Y        = [-2; -4; -6.0]
-        model    = PLS.fit(X,Y,nfactors=1,kernel="rbf",width=0.01)
-    	Y_pred     = PLS.predict(model,X)
+        model    = PLSRegressor.fit(X,Y,nfactors=1,kernel="rbf",width=0.01)
+    	Y_pred     = PLSRegressor.predict(model,X)
         @test mean(abs.(Y .- Y_pred)) < 1e-6
 
         X        = [1 2; 2 4; 4.0 6]
         Y        = [2; 4; 6.0]
-        model    = PLS.fit(X,Y,nfactors=1,kernel="rbf",width=0.01)
-    	Y_pred     = PLS.predict(model,X)
+        model    = PLSRegressor.fit(X,Y,nfactors=1,kernel="rbf",width=0.01)
+    	Y_pred     = PLSRegressor.predict(model,X)
         @test mean(abs.(Y .- Y_pred)) < 1e-6
 
     end
@@ -41,29 +41,29 @@
 
         X        = [1; 2; 3.0]
         Y        = [1 1; 2 2; 3 3.0]
-        model    = PLS.fit(X,Y,nfactors=1,kernel="rbf",width=0.01)
-    	Y_pred     = PLS.predict(model,X)
+        model    = PLSRegressor.fit(X,Y,nfactors=1,kernel="rbf",width=0.01)
+    	Y_pred     = PLSRegressor.predict(model,X)
         @test mean(abs.(Y .- Y_pred)) < 1e-6
 
         X        = [1; 2; 3.0]
         Y        = [1 -1; 2 -2; 3 -3.0]
-        model    = PLS.fit(X,Y,nfactors=1,kernel="rbf",width=0.01)
-    	Y_pred     = PLS.predict(model,X)
+        model    = PLSRegressor.fit(X,Y,nfactors=1,kernel="rbf",width=0.01)
+    	Y_pred     = PLSRegressor.predict(model,X)
         @test mean(abs.(Y .- Y_pred)) < 1e-6
 
         @testset "Linear Prediction Tests " begin
 
 
         X        = [1 2; 2 4; 4 6.0]
         Y        = [4 2;6 4;8 6.0]
-        model    = PLS.fit(X,Y,nfactors=1,kernel="rbf",width=0.01)
-    	Y_pred     = PLS.predict(model,X)
+        model    = PLSRegressor.fit(X,Y,nfactors=1,kernel="rbf",width=0.01)
+    	Y_pred     = PLSRegressor.predict(model,X)
         @test mean(abs.(Y .- Y_pred)) < 1e-6
 
         X           = [1 -2; 2 -4; 4 -6.0]
         Y           = [-4 -2;-6 -4;-8 -6.0]
-        model    = PLS.fit(X,Y,nfactors=1,kernel="rbf",width=0.01)
-    	Y_pred     = PLS.predict(model,X)
+        model    = PLSRegressor.fit(X,Y,nfactors=1,kernel="rbf",width=0.01)
+    	Y_pred     = PLSRegressor.predict(model,X)
         @test mean(abs.(Y .- Y_pred)) < 1e-6
 
 
@@ -82,14 +82,14 @@ end;
 	Xtr        = [1 -2; 2 -4; 4.0 -6]
 	Ytr        = [-2; -4; -6.0]
 	Xt         = [6 -8; 8 -10; 10.0 -12]
-	model1    = PLS.fit(Xtr,Ytr,nfactors=1,kernel="rbf",width=0.01)
-	pred1     = PLS.predict(model1,Xt)
+	model1    = PLSRegressor.fit(Xtr,Ytr,nfactors=1,kernel="rbf",width=0.01)
+	pred1     = PLSRegressor.predict(model1,Xt)
 
-	PLS.save(model1)
-	model2    = PLS.load()
+	PLSRegressor.save(model1)
+	model2    = PLSRegressor.load()
 
-	pred2     = PLS.predict(model2,Xt)
-    rm(PLS.MODEL_FILENAME)
+	pred2     = PLSRegressor.predict(model2,Xt)
+    rm(PLSRegressor.MODEL_FILENAME)
 	@test all(pred1 .== pred2)