-
-
Notifications
You must be signed in to change notification settings - Fork 1
Commit
This commit does not belong to any branch on this repository, and may belong to a fork outside of the repository.
- Loading branch information
1 parent
f587cca
commit 6b93028
Showing
2 changed files
with
373 additions
and
373 deletions.
There are no files selected for viewing
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| Original file line number | Diff line number | Diff line change |
|---|---|---|
| @@ -1,168 +1,168 @@ | ||
| data(iris) | ||
|
|
||
| X <- as.matrix(iris[,-5]) | ||
| y <- iris$Species | ||
|
|
||
| set.seed(123) | ||
| (index_train <- base::sample.int(n = nrow(X), | ||
| size = floor(0.8*nrow(X)), | ||
| replace = FALSE)) | ||
| X_train <- X[index_train, ] | ||
| y_train <- y[index_train] | ||
| X_test <- X[-index_train, ] | ||
| y_test <- y[-index_train] | ||
|
|
||
| obj1 <- learningmachine::Classifier$new(method = "lm") | ||
| obj2 <- learningmachine::Classifier$new(method = "ranger") | ||
| obj3 <- learningmachine::Classifier$new(method = "extratrees") | ||
| obj4 <- learningmachine::Classifier$new(method = "ridge") | ||
| obj5 <- learningmachine::Classifier$new(method = "bcn") | ||
| obj6 <- learningmachine::Classifier$new(method = "glmnet") | ||
| obj7 <- learningmachine::Classifier$new(method = "krr") | ||
| obj8 <- learningmachine::Classifier$new(method = "xgboost") | ||
| obj9 <- learningmachine::Classifier$new(method = "lm", nb_hidden = 3) | ||
| obj10 <- learningmachine::Classifier$new(method = "ranger", nb_hidden = 3) | ||
| obj11 <- learningmachine::Classifier$new(method = "extratrees", nb_hidden = 3) | ||
| obj12 <- learningmachine::Classifier$new(method = "ridge", nb_hidden = 3) | ||
| obj13 <- learningmachine::Classifier$new(method = "bcn", nb_hidden = 3) | ||
| obj14 <- learningmachine::Classifier$new(method = "glmnet", nb_hidden = 3) | ||
| obj15 <- learningmachine::Classifier$new(method = "krr", nb_hidden = 3) | ||
| obj16 <- learningmachine::Classifier$new(method = "xgboost", nb_hidden = 3) | ||
| obj17 <- learningmachine::Classifier$new(method = "ranger", | ||
| pi_method="kdesplitconformal") | ||
| obj18 <- learningmachine::Classifier$new(method = "ranger", | ||
| nb_hidden = 3, | ||
| pi_method="kdesplitconformal") | ||
|
|
||
| (obj1$get_type()) | ||
| (obj1$get_name()) | ||
| (obj1$get_method()) | ||
| (obj2$get_method()) | ||
| (obj3$get_method()) | ||
| (obj4$get_method()) | ||
| (obj5$get_method()) | ||
| (obj6$get_method()) | ||
| (obj7$get_method()) | ||
| (obj8$get_method()) | ||
|
|
||
| obj1$fit(X_train, y_train) | ||
| obj2$fit(X_train, y_train) | ||
| obj3$fit(X_train, y_train) | ||
| obj4$fit(X_train, y_train) | ||
| obj5$fit(X_train, y_train, | ||
| show_progress=FALSE) | ||
| obj6$fit(X_train, y_train, lambda=0.01) | ||
| obj7$fit(X_train, y_train) | ||
| obj8$fit(X_train, y_train, nrounds=10, verbose=FALSE) | ||
| obj9$fit(X_train, y_train) | ||
| obj10$fit(X_train, y_train) | ||
| obj11$fit(X_train, y_train) | ||
| obj12$fit(X_train, y_train) | ||
| obj13$fit(X_train, y_train, | ||
| show_progress=FALSE) | ||
| obj14$fit(X_train, y_train, lambda=0.01) | ||
| obj15$fit(X_train, y_train) | ||
| obj16$fit(X_train, y_train, nrounds=10, verbose=FALSE) | ||
| obj17$fit(X_train, y_train) | ||
| obj18$fit(X_train, y_train) | ||
|
|
||
|
|
||
|
|
||
| (acc1 <- mean((obj1$predict(X_test) == y_test))) | ||
| (acc2 <- mean((obj2$predict(X_test) == y_test))) | ||
| (acc3 <- mean((obj3$predict(X_test) == y_test))) | ||
| (acc4 <- mean((obj4$predict(X_test) == y_test))) | ||
| (acc5 <- mean((obj5$predict(X_test) == y_test))) | ||
| (acc6 <- mean((obj6$predict(X_test) == y_test))) | ||
| (acc7 <- mean((obj7$predict(X_test) == y_test))) | ||
| (acc8 <- mean((obj8$predict(X_test) == y_test))) | ||
| (acc9 <- mean((obj9$predict(X_test) == y_test))) | ||
| (acc10 <- mean((obj10$predict(X_test) == y_test))) | ||
| (acc11 <- mean((obj11$predict(X_test) == y_test))) | ||
| (acc12 <- mean((obj12$predict(X_test) == y_test))) | ||
| (acc13 <- mean((obj13$predict(X_test) == y_test))) | ||
| (acc14 <- mean((obj14$predict(X_test) == y_test))) | ||
| (acc15 <- mean((obj15$predict(X_test) == y_test))) | ||
| (acc16 <- mean((obj16$predict(X_test) == y_test))) | ||
|
|
||
| (prob1 <- obj1$predict_proba(X_test)[1, 1]) | ||
| (prob2 <- obj2$predict_proba(X_test)[1, 1]) | ||
| (prob3 <- obj3$predict_proba(X_test)[1, 1]) | ||
| (prob4 <- obj4$predict_proba(X_test)[1, 1]) | ||
| (prob5 <- obj5$predict_proba(X_test)[1, 1]) | ||
| (prob6 <- obj6$predict_proba(X_test)[1, 1]) | ||
| (prob7 <- obj7$predict_proba(X_test)[1, 1]) | ||
| (prob8 <- obj8$predict_proba(X_test)[1, 1]) | ||
| (prob9 <- obj9$predict_proba(X_test)[1, 1]) | ||
| (prob10 <- obj10$predict_proba(X_test)[1, 1]) | ||
| (prob11 <- obj11$predict_proba(X_test)[1, 1]) | ||
| (prob12 <- obj12$predict_proba(X_test)[1, 1]) | ||
| (prob13 <- obj13$predict_proba(X_test)[1, 1]) | ||
| (prob14 <- obj14$predict_proba(X_test)[1, 1]) | ||
| (prob15 <- obj15$predict_proba(X_test)[1, 1]) | ||
| (prob16 <- obj16$predict_proba(X_test)[1, 1]) | ||
|
|
||
| test_that("1 - checks on basic getters", { | ||
| expect_equal(obj1$get_type(), "classification") | ||
| expect_equal(obj1$get_name(), "Classifier") | ||
| expect_equal(obj1$get_method(), "lm") | ||
| expect_equal(obj2$get_method(), "ranger") | ||
| expect_equal(obj3$get_method(), "extratrees") | ||
| expect_equal(obj4$get_method(), "ridge") | ||
| expect_equal(obj5$get_method(), "bcn") | ||
| expect_equal(obj6$get_method(), "glmnet") | ||
| expect_equal(obj7$get_method(), "krr") | ||
| expect_equal(obj8$get_method(), "xgboost") | ||
| }) | ||
|
|
||
| test_that("2 - checks on basic fitting", { | ||
| expect_equal(round(acc1, 2), 0.77) | ||
| expect_equal(round(acc2, 2), 0.97) | ||
| expect_equal(round(acc3, 2), 0.97) | ||
| expect_equal(round(acc4, 2), 0.8) | ||
| expect_equal(round(acc5, 2), 0.17) | ||
| expect_equal(round(acc6, 2), 0.73) | ||
| expect_equal(round(acc7, 2), 0.97) | ||
| expect_equal(round(acc8, 2), 0.97) | ||
| }) | ||
|
|
||
| test_that("3 - basic checks on regression-based probas", { | ||
| expect_equal(as.numeric(round(prob1, 2)), 0.42) | ||
| expect_equal(as.numeric(round(prob2, 2)), 0.42) | ||
| expect_equal(as.numeric(round(prob3, 2)), 0.42) | ||
| expect_equal(as.numeric(round(prob4, 2)), 0.42) | ||
| expect_equal(as.numeric(round(prob5, 2)), 0.33) | ||
| expect_equal(as.numeric(round(prob6, 2)), 0.42) | ||
| expect_equal(as.numeric(round(prob7, 2)), 0.42) | ||
| expect_equal(as.numeric(round(prob8, 2)), 0.42) | ||
| }) | ||
|
|
||
|
|
||
| test_that("3 - checks on fitting qrns", { | ||
| expect_equal(round(acc9, 2), 0.77) | ||
| expect_equal(round(acc10, 2), 0.97) | ||
| expect_equal(round(acc11, 2), 0.97) | ||
| expect_equal(round(acc12, 2), 0.8) | ||
| expect_equal(round(acc13, 2), 0.17) | ||
| expect_equal(round(acc14, 2), 0.73) | ||
| expect_equal(round(acc15, 2), 0.97) | ||
| expect_equal(round(acc16, 2), 0.97) | ||
| }) | ||
|
|
||
| test_that("4 - checks on regression-based probas for qrns", { | ||
| expect_equal(as.numeric(round(prob9, 2)), 0.42) | ||
| expect_equal(as.numeric(round(prob10, 2)), 0.42) | ||
| expect_equal(as.numeric(round(prob11, 2)), 0.42) | ||
| expect_equal(as.numeric(round(prob12, 2)), 0.42) | ||
| expect_equal(as.numeric(round(prob13, 2)), 0.33) | ||
| expect_equal(as.numeric(round(prob14, 2)), 0.42) | ||
| expect_equal(as.numeric(round(prob15, 2)), 0.42) | ||
| expect_equal(as.numeric(round(prob16, 2)), 0.42) | ||
| }) | ||
|
|
||
| # test_that("5 - conformal no sims", { | ||
| # expect_equal(round(cv1, 2), 1) | ||
| # expect_equal(round(cv2, 2), 1) | ||
| # expect_equal(round(cv3, 2), 0.57) | ||
| # expect_equal(round(cv4, 2), 1) | ||
| # data(iris) | ||
|
|
||
| # X <- as.matrix(iris[,-5]) | ||
| # y <- iris$Species | ||
|
|
||
| # set.seed(123) | ||
| # (index_train <- base::sample.int(n = nrow(X), | ||
| # size = floor(0.8*nrow(X)), | ||
| # replace = FALSE)) | ||
| # X_train <- X[index_train, ] | ||
| # y_train <- y[index_train] | ||
| # X_test <- X[-index_train, ] | ||
| # y_test <- y[-index_train] | ||
|
|
||
| # obj1 <- learningmachine::Classifier$new(method = "lm") | ||
| # obj2 <- learningmachine::Classifier$new(method = "ranger") | ||
| # obj3 <- learningmachine::Classifier$new(method = "extratrees") | ||
| # obj4 <- learningmachine::Classifier$new(method = "ridge") | ||
| # obj5 <- learningmachine::Classifier$new(method = "bcn") | ||
| # obj6 <- learningmachine::Classifier$new(method = "glmnet") | ||
| # obj7 <- learningmachine::Classifier$new(method = "krr") | ||
| # obj8 <- learningmachine::Classifier$new(method = "xgboost") | ||
| # obj9 <- learningmachine::Classifier$new(method = "lm", nb_hidden = 3) | ||
| # obj10 <- learningmachine::Classifier$new(method = "ranger", nb_hidden = 3) | ||
| # obj11 <- learningmachine::Classifier$new(method = "extratrees", nb_hidden = 3) | ||
| # obj12 <- learningmachine::Classifier$new(method = "ridge", nb_hidden = 3) | ||
| # obj13 <- learningmachine::Classifier$new(method = "bcn", nb_hidden = 3) | ||
| # obj14 <- learningmachine::Classifier$new(method = "glmnet", nb_hidden = 3) | ||
| # obj15 <- learningmachine::Classifier$new(method = "krr", nb_hidden = 3) | ||
| # obj16 <- learningmachine::Classifier$new(method = "xgboost", nb_hidden = 3) | ||
| # obj17 <- learningmachine::Classifier$new(method = "ranger", | ||
| # pi_method="kdesplitconformal") | ||
| # obj18 <- learningmachine::Classifier$new(method = "ranger", | ||
| # nb_hidden = 3, | ||
| # pi_method="kdesplitconformal") | ||
|
|
||
| # (obj1$get_type()) | ||
| # (obj1$get_name()) | ||
| # (obj1$get_method()) | ||
| # (obj2$get_method()) | ||
| # (obj3$get_method()) | ||
| # (obj4$get_method()) | ||
| # (obj5$get_method()) | ||
| # (obj6$get_method()) | ||
| # (obj7$get_method()) | ||
| # (obj8$get_method()) | ||
|
|
||
| # obj1$fit(X_train, y_train) | ||
| # obj2$fit(X_train, y_train) | ||
| # obj3$fit(X_train, y_train) | ||
| # obj4$fit(X_train, y_train) | ||
| # obj5$fit(X_train, y_train, | ||
| # show_progress=FALSE) | ||
| # obj6$fit(X_train, y_train, lambda=0.01) | ||
| # obj7$fit(X_train, y_train) | ||
| # obj8$fit(X_train, y_train, nrounds=10, verbose=FALSE) | ||
| # obj9$fit(X_train, y_train) | ||
| # obj10$fit(X_train, y_train) | ||
| # obj11$fit(X_train, y_train) | ||
| # obj12$fit(X_train, y_train) | ||
| # obj13$fit(X_train, y_train, | ||
| # show_progress=FALSE) | ||
| # obj14$fit(X_train, y_train, lambda=0.01) | ||
| # obj15$fit(X_train, y_train) | ||
| # obj16$fit(X_train, y_train, nrounds=10, verbose=FALSE) | ||
| # obj17$fit(X_train, y_train) | ||
| # obj18$fit(X_train, y_train) | ||
|
|
||
|
|
||
|
|
||
| # (acc1 <- mean((obj1$predict(X_test) == y_test))) | ||
| # (acc2 <- mean((obj2$predict(X_test) == y_test))) | ||
| # (acc3 <- mean((obj3$predict(X_test) == y_test))) | ||
| # (acc4 <- mean((obj4$predict(X_test) == y_test))) | ||
| # (acc5 <- mean((obj5$predict(X_test) == y_test))) | ||
| # (acc6 <- mean((obj6$predict(X_test) == y_test))) | ||
| # (acc7 <- mean((obj7$predict(X_test) == y_test))) | ||
| # (acc8 <- mean((obj8$predict(X_test) == y_test))) | ||
| # (acc9 <- mean((obj9$predict(X_test) == y_test))) | ||
| # (acc10 <- mean((obj10$predict(X_test) == y_test))) | ||
| # (acc11 <- mean((obj11$predict(X_test) == y_test))) | ||
| # (acc12 <- mean((obj12$predict(X_test) == y_test))) | ||
| # (acc13 <- mean((obj13$predict(X_test) == y_test))) | ||
| # (acc14 <- mean((obj14$predict(X_test) == y_test))) | ||
| # (acc15 <- mean((obj15$predict(X_test) == y_test))) | ||
| # (acc16 <- mean((obj16$predict(X_test) == y_test))) | ||
|
|
||
| # (prob1 <- obj1$predict_proba(X_test)[1, 1]) | ||
| # (prob2 <- obj2$predict_proba(X_test)[1, 1]) | ||
| # (prob3 <- obj3$predict_proba(X_test)[1, 1]) | ||
| # (prob4 <- obj4$predict_proba(X_test)[1, 1]) | ||
| # (prob5 <- obj5$predict_proba(X_test)[1, 1]) | ||
| # (prob6 <- obj6$predict_proba(X_test)[1, 1]) | ||
| # (prob7 <- obj7$predict_proba(X_test)[1, 1]) | ||
| # (prob8 <- obj8$predict_proba(X_test)[1, 1]) | ||
| # (prob9 <- obj9$predict_proba(X_test)[1, 1]) | ||
| # (prob10 <- obj10$predict_proba(X_test)[1, 1]) | ||
| # (prob11 <- obj11$predict_proba(X_test)[1, 1]) | ||
| # (prob12 <- obj12$predict_proba(X_test)[1, 1]) | ||
| # (prob13 <- obj13$predict_proba(X_test)[1, 1]) | ||
| # (prob14 <- obj14$predict_proba(X_test)[1, 1]) | ||
| # (prob15 <- obj15$predict_proba(X_test)[1, 1]) | ||
| # (prob16 <- obj16$predict_proba(X_test)[1, 1]) | ||
|
|
||
| # test_that("1 - checks on basic getters", { | ||
| # expect_equal(obj1$get_type(), "classification") | ||
| # expect_equal(obj1$get_name(), "Classifier") | ||
| # expect_equal(obj1$get_method(), "lm") | ||
| # expect_equal(obj2$get_method(), "ranger") | ||
| # expect_equal(obj3$get_method(), "extratrees") | ||
| # expect_equal(obj4$get_method(), "ridge") | ||
| # expect_equal(obj5$get_method(), "bcn") | ||
| # expect_equal(obj6$get_method(), "glmnet") | ||
| # expect_equal(obj7$get_method(), "krr") | ||
| # expect_equal(obj8$get_method(), "xgboost") | ||
| # }) | ||
|
|
||
| # test_that("2 - checks on basic fitting", { | ||
| # expect_equal(round(acc1, 2), 0.77) | ||
| # expect_equal(round(acc2, 2), 0.97) | ||
| # expect_equal(round(acc3, 2), 0.97) | ||
| # expect_equal(round(acc4, 2), 0.8) | ||
| # expect_equal(round(acc5, 2), 0.17) | ||
| # expect_equal(round(acc6, 2), 0.73) | ||
| # expect_equal(round(acc7, 2), 0.97) | ||
| # expect_equal(round(acc8, 2), 0.97) | ||
| # }) | ||
|
|
||
| # test_that("3 - basic checks on regression-based probas", { | ||
| # expect_equal(as.numeric(round(prob1, 2)), 0.42) | ||
| # expect_equal(as.numeric(round(prob2, 2)), 0.42) | ||
| # expect_equal(as.numeric(round(prob3, 2)), 0.42) | ||
| # expect_equal(as.numeric(round(prob4, 2)), 0.42) | ||
| # expect_equal(as.numeric(round(prob5, 2)), 0.33) | ||
| # expect_equal(as.numeric(round(prob6, 2)), 0.42) | ||
| # expect_equal(as.numeric(round(prob7, 2)), 0.42) | ||
| # expect_equal(as.numeric(round(prob8, 2)), 0.42) | ||
| # }) | ||
|
|
||
|
|
||
| # test_that("3 - checks on fitting qrns", { | ||
| # expect_equal(round(acc9, 2), 0.77) | ||
| # expect_equal(round(acc10, 2), 0.97) | ||
| # expect_equal(round(acc11, 2), 0.97) | ||
| # expect_equal(round(acc12, 2), 0.8) | ||
| # expect_equal(round(acc13, 2), 0.17) | ||
| # expect_equal(round(acc14, 2), 0.73) | ||
| # expect_equal(round(acc15, 2), 0.97) | ||
| # expect_equal(round(acc16, 2), 0.97) | ||
| # }) | ||
|
|
||
| # test_that("4 - checks on regression-based probas for qrns", { | ||
| # expect_equal(as.numeric(round(prob9, 2)), 0.42) | ||
| # expect_equal(as.numeric(round(prob10, 2)), 0.42) | ||
| # expect_equal(as.numeric(round(prob11, 2)), 0.42) | ||
| # expect_equal(as.numeric(round(prob12, 2)), 0.42) | ||
| # expect_equal(as.numeric(round(prob13, 2)), 0.33) | ||
| # expect_equal(as.numeric(round(prob14, 2)), 0.42) | ||
| # expect_equal(as.numeric(round(prob15, 2)), 0.42) | ||
| # expect_equal(as.numeric(round(prob16, 2)), 0.42) | ||
| # }) | ||
|
|
||
| # # test_that("5 - conformal no sims", { | ||
| # # expect_equal(round(cv1, 2), 1) | ||
| # # expect_equal(round(cv2, 2), 1) | ||
| # # expect_equal(round(cv3, 2), 0.57) | ||
| # # expect_equal(round(cv4, 2), 1) | ||
| # # }) |
Oops, something went wrong.