Merge pull request #77 from ACEsuit/sklearn-update

Update for sklearn 1.3.

.github/workflows/CI.yml

@@ -19,6 +19,7 @@ jobs:
       matrix:
         version:
           - '1.9'
+          - '1.10'
           - '1'
           - 'nightly'
         python-version: 

Project.toml

@@ -30,7 +30,7 @@ julia = "1.9"
 IterativeSolvers = "0.9.2"
 MLJ = "0.19"
 MLJLinearModels = "0.9"
-MLJScikitLearnInterface = "0.5"
+MLJScikitLearnInterface = "0.7"
 LowRankApprox = "0.5.3"
 Optim = "1.7"
 ParallelDataTransfer = "0.5.0"

docs/src/index.md

@@ -34,7 +34,7 @@ ARDRegressor = @load ARDRegressor pkg=MLJScikitLearnInterface
 
 # Create the solver itself and give it parameters
 solver = ARDRegressor(
-    n_iter = 300,
+    max_iter = 300,
     tol = 1e-3,
     threshold_lambda = 10000
 )

ext/ACEfit_MLJScikitLearnInterface_ext.jl

@@ -21,7 +21,7 @@ ARDRegressor = @load ARDRegressor pkg=MLJScikitLearnInterface
 
 # Create the solver itself and give it parameters
 solver = ARDRegressor(
-    n_iter = 300,
+    max_iter = 300,
     tol = 1e-3,
     threshold_lambda = 10000
     # more params
@@ -43,4 +43,4 @@ function  ACEfit.solve(solver, A, y)
     return Dict{String, Any}("C" => pyconvert(Array, c) )
 end
 
-end
+end

ext/ACEfit_PythonCall_ext.jl

@@ -7,13 +7,13 @@ using PythonCall
 function ACEfit.solve(solver::ACEfit.SKLEARN_BRR, A, y)
     @info "Entering SKLEARN_BRR"
     BRR = pyimport("sklearn.linear_model")."BayesianRidge"
-    clf = BRR(n_iter = solver.n_iter, tol = solver.tol, fit_intercept = true,
+    clf = BRR(max_iter = solver.max_iter, tol = solver.tol, fit_intercept = true,
               compute_score = true)
     clf.fit(A, y)
-    if length(clf.scores_) < solver.n_iter
+    if length(clf.scores_) < solver.max_iter
         @info "BRR converged to tol=$(solver.tol) after $(length(clf.scores_)) iterations."
     else
-        @warn "\nBRR did not converge to tol=$(solver.tol) after n_iter=$(solver.n_iter) iterations.\n"
+        @warn "\nBRR did not converge to tol=$(solver.tol) after max_iter=$(solver.max_iter) iterations.\n"
     end
     c = clf.coef_
     return Dict{String, Any}("C" => pyconvert(Array, c) )
@@ -22,17 +22,17 @@ end
 
 function ACEfit.solve(solver::ACEfit.SKLEARN_ARD, A, y)
     ARD = pyimport("sklearn.linear_model")."ARDRegression"
-    clf = ARD(n_iter = solver.n_iter, threshold_lambda = solver.threshold_lambda,
+    clf = ARD(max_iter = solver.max_iter, threshold_lambda = solver.threshold_lambda,
               tol = solver.tol,
               fit_intercept = true, compute_score = true)
     clf.fit(A, y)
-    if length(clf.scores_) < solver.n_iter
+    if length(clf.scores_) < solver.max_iter
         @info "ARD converged to tol=$(solver.tol) after $(length(clf.scores_)) iterations."
     else
-        @warn "\n\nARD did not converge to tol=$(solver.tol) after n_iter=$(solver.n_iter) iterations.\n\n"
+        @warn "\n\nARD did not converge to tol=$(solver.tol) after max_iter=$(solver.max_iter) iterations.\n\n"
     end
     c = clf.coef_
     return Dict{String, Any}("C" => pyconvert(Array,c) )
 end
 
-end
+end

src/solvers.jl

@@ -124,12 +124,12 @@ SKLEARN_BRR
 """
 struct SKLEARN_BRR
     tol::Number
-    n_iter::Integer
+    max_iter::Integer
 end
 
-function SKLEARN_BRR(; tol = 1e-3, n_iter = 300)
+function SKLEARN_BRR(; tol = 1e-3, max_iter = 300)
     @warn "SKLearn will transition to MLJ in future, please upgrade your script to reflect this."
-    SKLEARN_BRR(tol, n_iter)
+    SKLEARN_BRR(tol, max_iter)
 end
 
 # solve(solver::SKLEARN_BRR, ...) is implemented in ext/
@@ -138,14 +138,14 @@ end
 SKLEARN_ARD
 """
 struct SKLEARN_ARD
-    n_iter::Integer
+    max_iter::Integer
     tol::Number
     threshold_lambda::Number
 end
 
-function SKLEARN_ARD(; n_iter = 300, tol = 1e-3, threshold_lambda = 10000)
+function SKLEARN_ARD(; max_iter = 300, tol = 1e-3, threshold_lambda = 10000)
     @warn "SKLearn will transition to MLJ in future, please upgrade your script to reflect this."
-    SKLEARN_ARD(n_iter, tol, threshold_lambda)
+    SKLEARN_ARD(max_iter, tol, threshold_lambda)
 end
 
 # solve(solver::SKLEARN_ARD, ...) is implemented in ext/

test/test_mlj.jl

@@ -32,11 +32,11 @@ C = results["C"]
 @info(" ... MLJ SKLearn ARD")
 ARDRegressor = @load ARDRegressor pkg=MLJScikitLearnInterface
 solver = ARDRegressor(
-    n_iter = 300,
+    max_iter = 300,
     tol = 1e-3,
     threshold_lambda = 10000
 )
 results = ACEfit.solve(solver, A, y)
 C = results["C"]
 @show norm(A * C - y)
-@show norm(C)
+@show norm(C)