Merge remote-tracking branch 'origin/main' into mm/workflows

Author: MatthewMiddlehurst

.github/utilities/ai_spam.py

@@ -0,0 +1,27 @@
+"""Script for handling AI Spam label on pull requests.
+
+Triggered when AI Spam label is added to a PR,
+it adds a comment and closes the PR.
+"""
+
+import json
+import os
+
+from github import Github
+
+context_dict = json.loads(os.getenv("CONTEXT_GITHUB"))
+
+repo_name = context_dict["repository"]
+g = Github(os.getenv("GITHUB_TOKEN"))
+repo = g.get_repo(repo_name)
+pr_number = context_dict["event"]["pull_request"]["number"]
+pr = repo.get_pull(pr_number)
+label_name = context_dict["event"]["label"]["name"]
+
+if label_name == "AI Spam":
+    comment_body = (
+        "This pull request has been flagged with the **AI Spam** label.\n\n"
+        "This PR is being closed."
+    )
+    pr.create_issue_comment(comment_body)
+    pr.edit(state="closed")

.github/utilities/issue_assign.py

@@ -1,12 +1,8 @@
 """Script for the GitHub issue self-assign bot.
 
-It checks if a comment on an issue or PR includes the trigger
-phrase (as defined) and a mentioned user.
-If it does, it assigns the issue to the mentioned user.
-Users without write access can only have up to 2 open issues assigned.
-Users with write access (or admin) are exempt from this limit.
-If a non-write user already has 2 or more open issues, the bot
-comments on the issue with links to the currently assigned open issues.
+Checks if a comment on an issue or PR includes the trigger phrase and a mentioned user.
+If it does, it assigns or unassigns the issue to the mentioned user if they have
+permissions.
 """
 
 import json
@@ -22,54 +18,75 @@
 repo = g.get_repo(repo)
 issue_number = context_dict["event"]["issue"]["number"]
 issue = repo.get_issue(number=issue_number)
-comment_body = context_dict["event"]["comment"]["body"]
+issue_labels = {label.name.lower() for label in issue.labels}
 pr = context_dict["event"]["issue"].get("pull_request")
+comment_body = context_dict["event"]["comment"]["body"]
 commenter = context_dict["event"]["comment"]["user"]["login"]
+commenter_permission = repo.get_collaborator_permission(commenter)
+has_write_permission = commenter_permission not in ["admin", "write"]
+
+restricted_labels = {"meta-issue"}
 
+# Assign tagged used to the issue if the comment includes the trigger phrase
 body = comment_body.lower()
-if "@aeon-actions-bot" in body and not pr:
-    # Assign commenter if comment includes "assign me"
-    if "assign me" in body:
-        issue.add_to_assignees(commenter)
-    # Assign tagged used to the issue if the comment includes the trigger phrase
-    elif "assign" in body:
+if "@aeon-actions-bot" in body and "assign" in body and not pr:
+    # Check if the issue has any restricted labels for auto assignment
+    label_intersect = issue_labels & restricted_labels
+    if len(label_intersect) > 0:
+        issue.create_comment(
+            f"This issue contains the following restricted label(s): "
+            f"{', '.join(label_intersect)}. Cannot assign to users."
+        )
+    else:
+        # collect any mentioned (@username) users
         mentioned_users = re.findall(r"@[a-zA-Z0-9_-]+", comment_body)
         mentioned_users = [user[1:] for user in mentioned_users]
         mentioned_users.remove("aeon-actions-bot")
+        # Assign commenter if comment includes "assign me"
+        if "assign me" in body:
+            mentioned_users.append(commenter)
+        mentioned_users = set(mentioned_users)
 
+        access_error = False
         for user in mentioned_users:
-            user_obj = g.get_user(user)
-            permission = repo.get_collaborator_permission(user_obj)
+            # Can only assign others if the commenter has write access
+            if user != commenter and not has_write_permission:
+                if not access_error:
+                    issue.create_comment(
+                        "Cannot assign other users to issues without write access."
+                    )
+                    access_error = True
+                continue
+
+            # If the user is already assigned to this issue, remove them
+            if user in [assignee.login for assignee in issue.assignees]:
+                issue.remove_from_assignees(user)
+                continue
 
-            if permission in ["admin", "write"]:
+            # If the commenter has write access, just assign
+            if has_write_permission:
                 issue.add_to_assignees(user)
             else:
-                # First check if the user is already assigned to this issue
-                if user in [assignee.login for assignee in issue.assignees]:
-                    continue
-
                 # search for open issues only
                 query = f"repo:{repo.full_name} is:issue is:open assignee:{user}"
                 issues_assigned_to_user = g.search_issues(query)
                 assigned_count = issues_assigned_to_user.totalCount
 
-                if assigned_count >= 2:
+                if assigned_count >= 3:
                     # link to issue
                     assigned_issues_list = [
                         f"[#{assigned_issue.number}]({assigned_issue.html_url})"
                         for assigned_issue in issues_assigned_to_user
                     ]
 
-                    comment_message = (
-                        f"@{user}, you already have {assigned_count} "
-                        f"open issues assigned."
-                        "Users without write access are limited to self-assigning two"
-                        "issues.\n\n"
-                        "Here are the open issues assigned to you:\n"
+                    issue.create_comment(
+                        f"@{user}, already has {assigned_count} open issues assigned."
+                        "Users without write access are limited to self-assigning "
+                        "three issues.\n\n"
+                        "Here are the open issues assigned:\n"
                         + "\n".join(
                             f"- {issue_link}" for issue_link in assigned_issues_list
                         )
                     )
-                    issue.create_comment(comment_message)
                 else:
                     issue.add_to_assignees(user)

.github/workflows/ai_spam.yml

@@ -0,0 +1,42 @@
+name: AI Spam Detection On PR
+
+on:
+  pull_request:
+    types: [labeled]
+
+concurrency:
+  group: ${{ github.workflow }}-${{ github.event.pull_request.number }}
+  cancel-in-progress: true
+
+jobs:
+  ai-spam-present:
+    if: ${{ github.event.label.name == 'AI Spam' }}
+    runs-on: ubuntu-24.04
+
+    steps:
+      - name: Create app token
+        uses: actions/create-github-app-token@v1
+        id: app-token
+        with:
+          app-id: ${{ vars.PR_APP_ID }}
+          private-key: ${{ secrets.PR_APP_KEY }}
+
+      - name: Checkout main
+        uses: actions/checkout@v4
+        with:
+          sparse-checkout: .github/utilities
+
+      - name: Setup Python 3.11
+        uses: actions/setup-python@v5
+        with:
+          python-version: "3.11"
+
+      - name: Install PyGithub
+        run: pip install -Uq PyGithub
+
+      - name: Process AI Spam
+        id: handle_spam
+        run: python .github/utilities/ai_spam.py
+        env:
+          CONTEXT_GITHUB: ${{ toJson(github) }}
+          GITHUB_TOKEN: ${{ steps.app-token.outputs.token }}

aeon/anomaly_detection/series/distance_based/_rockad.py

@@ -176,6 +176,8 @@ def _inner_fit(self, X: np.ndarray) -> None:
 
         if self.power_transform:
             self.power_transformer_ = PowerTransformer()
+            # todo check if this is still an issue with scikit-learn >= 1.7.0
+            # when lower bound is raised
             try:
                 Xtp = self.power_transformer_.fit_transform(Xt)
 

aeon/anomaly_detection/series/distance_based/tests/test_rockad.py

@@ -73,11 +73,6 @@ def test_rockad_incorrect_input():
     ):
         ad = ROCKAD(stride=1, window_size=100)
         ad.fit(train_series)
-    with pytest.warns(
-        UserWarning, match=r"Power Transform failed and thus has been disabled."
-    ):
-        ad = ROCKAD(stride=1, window_size=5)
-        ad.fit(train_series)
     with pytest.raises(
         ValueError, match=r"window shape cannot be larger than input array shape"
     ):

aeon/classification/sklearn/_continuous_interval_tree.py

@@ -21,6 +21,7 @@
 from sklearn.exceptions import NotFittedError
 from sklearn.utils import check_random_state
 from sklearn.utils.multiclass import check_classification_targets
+from sklearn.utils.validation import validate_data
 
 
 class _TreeNode:
@@ -374,7 +375,8 @@ def fit(self, X, y):
         """
         # data processing
         X = self._check_X(X)
-        X, y = self._validate_data(
+        X, y = validate_data(
+            self,
             X=X,
             y=y,
             ensure_min_samples=2,
@@ -464,8 +466,8 @@ def predict_proba(self, X):
 
         # data processing
         X = self._check_X(X)
-        X = self._validate_data(
-            X=X, reset=False, force_all_finite="allow-nan", accept_sparse=False
+        X = validate_data(
+            self, X=X, reset=False, force_all_finite="allow-nan", accept_sparse=False
         )
 
         dists = np.zeros((X.shape[0], self.n_classes_))

aeon/classification/sklearn/_rotation_forest_classifier.py

@@ -19,6 +19,7 @@
 from sklearn.tree import DecisionTreeClassifier
 from sklearn.utils import check_random_state
 from sklearn.utils.multiclass import check_classification_targets
+from sklearn.utils.validation import validate_data
 
 from aeon.base._base import _clone_estimator
 from aeon.utils.validation import check_n_jobs
@@ -192,7 +193,7 @@ def predict_proba(self, X) -> np.ndarray:
 
         # data processing
         X = self._check_X(X)
-        X = self._validate_data(X=X, reset=False, accept_sparse=False)
+        X = validate_data(self, X=X, reset=False, accept_sparse=False)
 
         # replace missing values with 0 and remove useless attributes
         X = X[:, self._useful_atts]
@@ -299,7 +300,7 @@ def fit_predict_proba(self, X, y) -> np.ndarray:
     def _fit_rotf(self, X, y, save_transformed_data: bool = False):
         # data processing
         X = self._check_X(X)
-        X, y = self._validate_data(X=X, y=y, ensure_min_samples=2, accept_sparse=False)
+        X, y = validate_data(self, X=X, y=y, ensure_min_samples=2, accept_sparse=False)
         check_classification_targets(y)
 
         self._n_jobs = check_n_jobs(self.n_jobs)

aeon/regression/sklearn/_rotation_forest_regressor.py

@@ -19,6 +19,7 @@
 from sklearn.exceptions import NotFittedError
 from sklearn.tree import DecisionTreeRegressor
 from sklearn.utils import check_random_state
+from sklearn.utils.validation import validate_data
 
 from aeon.base._base import _clone_estimator
 from aeon.utils.validation import check_n_jobs
@@ -168,7 +169,7 @@ def predict(self, X) -> np.ndarray:
 
         # data processing
         X = self._check_X(X)
-        X = self._validate_data(X=X, reset=False, accept_sparse=False)
+        X = validate_data(self, X=X, reset=False, accept_sparse=False)
 
         # replace missing values with 0 and remove useless attributes
         X = X[:, self._useful_atts]
@@ -222,7 +223,7 @@ def fit_predict(self, X, y) -> np.ndarray:
     def _fit_rotf(self, X, y, save_transformed_data: bool = False):
         # data processing
         X = self._check_X(X)
-        X, y = self._validate_data(X=X, y=y, ensure_min_samples=2, accept_sparse=False)
+        X, y = validate_data(self, X=X, y=y, ensure_min_samples=2, accept_sparse=False)
 
         self._label_average = np.mean(y)
 

pyproject.toml

@@ -50,7 +50,7 @@ dependencies = [
     "numpy>=1.21.0,<2.3.0",
     "packaging>=20.0",
     "pandas>=2.0.0,<2.4.0",
-    "scikit-learn>=1.0.0,<1.7.0",
+    "scikit-learn>=1.0.0,<1.8.0",
     "scipy>=1.9.0,<1.16.0",
     "typing-extensions>=4.6.0",
 ]