Pkg_restructure (#20)

* 🎨 move to src layout

* 🔥 moved to pyproject.toml (from setup.py)

* 🎨🔖 switch to dynamic versioning based on tags

- each commit gets his own version relativ to the last tag
- releases do not require manuel setting of version (they pick it up from the git tags)

* 🐛 set src pkg path correctly for readthedocs

* 🐛 forgot to activate scm via setting an entry in pyproject.toml

see https://setuptools-scm.readthedocs.io

docs/README.md

@@ -25,7 +25,7 @@ Options:
 ```bash	
 # pwd: docs
 # apidoc
-sphinx-apidoc --force --implicit-namespaces --module-first -o reference ../acore
+sphinx-apidoc --force --implicit-namespaces --module-first -o reference ../src/acore
 # build docs
 sphinx-build -n -W --keep-going -b html ./ ./_build/
 ```

docs/conf.py

@@ -223,7 +223,7 @@
     from pathlib import Path
 
     PROJECT_ROOT = Path(__file__).parent.parent
-    PACKAGE_ROOT = PROJECT_ROOT / "acore"
+    PACKAGE_ROOT = PROJECT_ROOT / "src" / "acore"
 
     def run_apidoc(_):
         from sphinx.ext import apidoc

pyproject.toml

@@ -2,7 +2,7 @@
 authors = [{ name = "Alberto Santos Delgado", email = "[email protected]" },
            { name = "Henry Webel", email = "[email protected]"}]
 name = "acore"
-version = "0.1.2"
+dynamic = ["version"]
 description = "A Python package with statistical functions to analyse multimodal molecular data"
 license = { text = "GNU General Public License v3" }
 readme = "README.rst"
@@ -68,9 +68,11 @@ Documentation = "https://analytics-core.readthedocs.io/"
 
 
 [build-system]
-requires = ["setuptools"]
+requires = ["setuptools>=64", "setuptools_scm>=8"]
 build-backend = "setuptools.build_meta"
 
+[tool.setuptools_scm]
+
 # [entrypoints]
 # console_scripts = {
 #     "acore = acore.cli:main"

setup.py

@@ -1,18 +1,3 @@
-#!/usr/bin/env python
+from setuptools import setup
 
-"""The setup script."""
-
-from setuptools import setup, find_packages
-
-setup(
-    entry_points={
-        'console_scripts': [
-            'acore=acore.cli:main',
-        ],
-    },
-    include_package_data=True,
-    keywords='acore',
-    packages=find_packages(include=['acore', 'acore.*']),
-    url='https://github.com/Multiomics-Analytics-Group/acore',
-    zip_safe=False,
-)
+setup()

src/acore/plotting/metrics.py

@@ -0,0 +1,81 @@
+"""Plot metrics for binary classification."""
+
+import matplotlib
+import matplotlib.pyplot as plt
+import pandas as pd
+from njab.sklearn.types import Results, ResultsSplit
+
+LIMITS = (-0.05, 1.05)
+
+
+def plot_split_auc(
+    result: ResultsSplit, name: str, ax: matplotlib.axes.Axes
+) -> matplotlib.axes.Axes:
+    """Add receiver operation curve to ax of a split of the data."""
+    col_name = f"{name} (auc: {result.auc:.3f})"
+    roc = pd.DataFrame(result.roc, index="fpr tpr cutoffs".split()).rename(
+        {"tpr": col_name}
+    )
+    ax = roc.T.plot(
+        "fpr",
+        col_name,
+        xlabel="false positive rate",
+        ylabel="true positive rate",
+        style=".-",
+        ylim=LIMITS,
+        xlim=LIMITS,
+        ax=ax,
+    )
+    return ax
+
+
+# ! should be roc
+def plot_auc(
+    results: Results,
+    ax: matplotlib.axes.Axes = None,
+    label_train="train",
+    label_test="test",
+    **kwargs,
+) -> matplotlib.axes.Axes:
+    """Plot ROC curve for train and test data."""
+    if ax is None:
+        fig, ax = plt.subplots(1, 1, **kwargs)
+    ax = plot_split_auc(results.train, f"{label_train}", ax)
+    ax = plot_split_auc(results.test, f"{label_test}", ax)
+    return ax
+
+
+def plot_split_prc(
+    result: ResultsSplit, name: str, ax: matplotlib.axes.Axes
+) -> matplotlib.axes.Axes:
+    """Add precision recall curve to ax of a split of the data."""
+    col_name = f"{name} (aps: {result.aps:.3f})"
+    roc = pd.DataFrame(result.prc, index="precision recall cutoffs".split()).rename(
+        {"precision": col_name}
+    )
+    ax = roc.T.plot(
+        "recall",
+        col_name,
+        xlabel="true positive rate",
+        ylabel="precision",
+        style=".-",
+        ylim=LIMITS,
+        xlim=LIMITS,
+        ax=ax,
+    )
+    return ax
+
+
+def plot_prc(
+    results: ResultsSplit,
+    ax: matplotlib.axes.Axes = None,
+    label_train="train",
+    label_test="test",
+    **kwargs,
+):
+    """Plot precision recall curve for train and test data."""
+    if ax is None:
+        fig, ax = plt.subplots(1, 1, **kwargs)
+    ax = plot_split_prc(results.train, f"{label_train}", ax)
+    ax = plot_split_prc(results.test, f"{label_test}", ax)
+    return ax