Merge branch 'sktime:main' into main

sktime · May 20, 2024 · 3c619d2 · 3c619d2
2 parents 53d4171 + e0eb67c
commit 3c619d2
Show file tree

Hide file tree

Showing 82 changed files with 9,151 additions and 2,006 deletions.
diff --git a/.all-contributorsrc b/.all-contributorsrc
@@ -175,6 +175,15 @@
       "contributions": [
         "maintenance"
       ]
+    },
+    {
+      "login": "szepeviktor",
+      "name": "Viktor Szépe",
+      "avatar_url": "https://avatars.githubusercontent.com/u/952007?v=4",
+      "profile": "https://github.com/szepeviktor",
+      "contributions": [
+        "doc"
+      ]
     }
   ]
 }
diff --git a/.binder/Dockerfile b/.binder/Dockerfile
@@ -0,0 +1,22 @@
+# This Dockerfile is used to build skpro when launching binder.
+# Find out more at: https://mybinder.readthedocs.io/en/latest/index.html
+
+FROM  jupyter/scipy-notebook:python-3.11.6
+# Set up user to avoid running as root
+ARG NB_USER
+ARG NB_UID
+ENV USER ${NB_USER}
+ENV HOME /home/${NB_USER}
+
+# Binder will automatically clone the repo, but we need to make sure the
+# contents of our repo are in the ${HOME} directory
+COPY . ${HOME}
+USER root
+RUN chown -R ${NB_UID} ${HOME}
+
+# Switch user and directory
+USER ${USER}
+WORKDIR ${HOME}
+
+# Install extra requirements and skpro based on main branch
+RUN pip install --upgrade pip --no-cache-dir && pip install .[binder]
diff --git a/README.md b/README.md
@@ -1,6 +1,6 @@
 <a href="https://skpro.readthedocs.io/en/latest"><img src="https://github.com/sktime/skpro/blob/main/docs/source/images/skpro-banner.png" width="500" align="right" /></a>
 
-:rocket: **Version 2.2.2 out now!** [Read the release notes here.](https://skpro.readthedocs.io/en/latest/changelog.html).
+:rocket: **Version 2.3.0 out now!** [Read the release notes here.](https://skpro.readthedocs.io/en/latest/changelog.html).
 
 `skpro` is a library for supervised probabilistic prediction in python.
 It provides `scikit-learn`-like, `scikit-base` compatible interfaces to:
@@ -59,7 +59,7 @@ We strongly believe in the value of sharing help publicly, as it allows a wider
 | :woman_technologist: **Usage Questions**          | [GitHub Discussions] · [Stack Overflow] |
 | :speech_balloon: **General Discussion**        | [GitHub Discussions] |
 | :factory: **Contribution & Development** | `dev-chat` channel · [Discord] |
-| :globe_with_meridians: **Community collaboration session** | [Discord] - Fridays 4 pm UTC, dev/meet-ups channel |
+| :globe_with_meridians: **Community collaboration session** | [Discord] - Fridays 13 UTC, dev/meet-ups channel |
 
 [github issue tracker]: https://github.com/sktime/skpro/issues
 [github discussions]: https://github.com/sktime/skpro/discussions
@@ -89,7 +89,7 @@ and for interfacing popular second- and third-party components, such as [cyclic-
 | **[Probabilistic tabular regression]** | maturing | [Tutorial](https://github.com/sktime/skpro/blob/main/examples/01_skpro_intro.ipynb) · [API Reference](https://skpro.readthedocs.io/en/latest/api_reference/regression.html) · [Extension Template](https://github.com/sktime/skpro/blob/main/extension_templates/regression.py) |
 | **[Time-to-event (survival) prediction]** | experimental | [API Reference](https://skpro.readthedocs.io/en/latest/api_reference/survival.html) · [Extension Template](https://github.com/sktime/skpro/blob/main/extension_templates/survival.py) |
 | **[Performance metrics]** | maturing | [API Reference](https://skpro.readthedocs.io/en/latest/api_reference/metrics.html) |
-| **[Probability distributions]** | maturing | [API Reference](https://skpro.readthedocs.io/en/latest/api_reference/distributions.html) |
+| **[Probability distributions]** | maturing | [Tutorial](https://github.com/sktime/skpro/blob/main/examples/03_skpro_distributions.ipynb) · [API Reference](https://skpro.readthedocs.io/en/latest/api_reference/distributions.html) · [Extension Template](https://github.com/sktime/skpro/blob/main/extension_templates/distributions.py) |
 
 [Probabilistic tabular regression]: https://github.com/sktime/skpro/tree/main/skpro/regression
 [Time-to-event (survival) prediction]: https://github.com/sktime/skpro/tree/main/skpro/survival

diff --git a/build_tools/changelog.py b/build_tools/changelog.py
@@ -4,23 +4,22 @@
 from collections import defaultdict
 from typing import Dict, List
 
-import httpx
-from dateutil import parser
-
 HEADERS = {
     "Accept": "application/vnd.github.v3+json",
 }
 
 if os.getenv("GITHUB_TOKEN") is not None:
     HEADERS["Authorization"] = f"token {os.getenv('GITHUB_TOKEN')}"
 
-OWNER = "skpro"
+OWNER = "sktime"
 REPO = "skpro"
 GITHUB_REPOS = "https://api.github.com/repos"
 
 
 def fetch_merged_pull_requests(page: int = 1) -> List[Dict]:  # noqa
     "Fetch a page of pull requests"
+    import httpx
+
     params = {
         "base": "main",
         "state": "closed",
@@ -38,6 +37,8 @@ def fetch_merged_pull_requests(page: int = 1) -> List[Dict]:  # noqa
 
 
 def fetch_latest_release():  # noqa
+    import httpx
+
     response = httpx.get(
         f"{GITHUB_REPOS}/{OWNER}/{REPO}/releases/latest", headers=HEADERS
     )
@@ -50,6 +51,7 @@ def fetch_latest_release():  # noqa
 
 def fetch_pull_requests_since_last_release() -> List[Dict]:  # noqa
     "Fetch pull requests and filter based on merged date"
+    from dateutil import parser
 
     release = fetch_latest_release()
     published_at = parser.parse(release["published_at"])
@@ -72,6 +74,8 @@ def fetch_pull_requests_since_last_release() -> List[Dict]:  # noqa
 
 def github_compare_tags(tag_left: str, tag_right: str = "HEAD"):  # noqa
     "Compare commit between two tags"
+    import httpx
+
     response = httpx.get(
         f"{GITHUB_REPOS}/{OWNER}/{REPO}/compare/{tag_left}...{tag_right}"
     )
@@ -128,6 +132,8 @@ def render_row(pr):  # noqa
 def render_changelog(prs, assigned):  # noqa
     # sourcery skip: use-named-expression
     "Render changelog"
+    from dateutil import parser
+
     for title, _ in assigned.items():
         pr_group = [prs[i] for i in assigned[title]]
         if pr_group:

diff --git a/docs/introduction.rst b/docs/introduction.rst
@@ -33,7 +33,7 @@ Let's have a look at an example of Boston Housing price prediction (using sklear
 
 If you are familiar with scikit-learn you will recognise that we define and train a model on the boston housing dataset and obtain the test prediction ``y_pred``. Furthermore, we use a loss function to calculate the loss between the predicted points and the true values -- nothing unexpected there.
 
-Crucially, however, the skpro model does not just return a list of numbers or point predictions here. Instead, ``y_pred`` is a probablistic prediction, i.e. it represents probability distributions for each individual data point.
+Crucially, however, the skpro model does not just return a list of numbers or point predictions here. Instead, ``y_pred`` is a probabilistic prediction, i.e. it represents probability distributions for each individual data point.
 We can, for instance, obtain the standard deviation of the predicted distribution that corresponds with the first (0th) test point (or any other test point distribution) ::
 
     >>> y_pred[0].std()
@@ -44,7 +44,7 @@ Notably, the interface represents all distributional properties, including the d
     >>> y_pred[0].pdf(x=42)
     0.00192808538756
 
-Furthermore, it is possible to conveniently access the distributional properties in their vectorized form accross the test sample::
+Furthermore, it is possible to conveniently access the distributional properties in their vectorized form across the test sample::
 
     >>> y_pred.std().shape
     (152,)

diff --git a/docs/source/_static/switcher.json b/docs/source/_static/switcher.json
@@ -5,7 +5,12 @@
     "url": "https://skpro.readthedocs.io/en/latest/"
   },
   {
-    "name": "2.2.2 (stable)",
+    "name": "2.3.0 (stable)",
+    "version": "stable",
+    "url": "https://skpro.readthedocs.io/en/v2.3.0/"
+  },
+  {
+    "name": "2.2.2",
     "version": "stable",
     "url": "https://skpro.readthedocs.io/en/v2.2.2/"
   },

diff --git a/docs/source/api_reference/distributions.rst b/docs/source/api_reference/distributions.rst
@@ -35,7 +35,9 @@ Continuous support
     :toctree: auto_generated/
     :template: class.rst
 
+    Beta
     ChiSquared
+    Exponential
     Fisk
     Laplace
     Logistic
@@ -65,8 +67,10 @@ Non-parametric and empirical distributions
     :toctree: auto_generated/
     :template: class.rst
 
+    Delta
     Empirical
     QPD_Empirical
+    QPD_Johnson
     QPD_U
     QPD_S
     QPD_B

diff --git a/docs/source/api_reference/regression.rst b/docs/source/api_reference/regression.rst
@@ -82,12 +82,13 @@ take one or multiple ``sklearn`` estimators and adda probabilistic prediction mo
 
     MapieRegressor
 
-.. currentmodule:: skpro.regression.gradient_boosting
+.. currentmodule:: skpro.regression.ensemble
 
 .. autosummary::
     :toctree: auto_generated/
     :template: class.rst
 
+    BaggingRegressor
     NGBoostRegressor
 
 .. currentmodule:: skpro.regression.cyclic_boosting
@@ -98,6 +99,21 @@ take one or multiple ``sklearn`` estimators and adda probabilistic prediction mo
 
     CyclicBoosting
 
+
+Naive regressors and baselines
+------------------------------
+
+This section lists simple regressors which can be used as baselines.
+
+.. currentmodule:: skpro.regression.delta
+
+.. autosummary::
+    :toctree: auto_generated/
+    :template: class.rst
+
+    DeltaPointRegressor
+
+
 Linear regression
 -----------------
 

diff --git a/docs/source/api_reference/survival.rst b/docs/source/api_reference/survival.rst
@@ -138,6 +138,7 @@ Tree ensemble models
     SurvivalForestXtraSkSurv
     SurvGradBoostSkSurv
     SurvGradBoostCompSkSurv
+    NGBoostSurvival
 
 Base
 ----