Merge pull request #562 from ICB-DCM/develop

Release 0.12.3

Author: yannikschaelte

.github/workflows/ci.yml

@@ -45,7 +45,7 @@ jobs:
       run: tox -e visualization
 
     - name: Coverage
-      uses: codecov/codecov-action@v1
+      uses: codecov/codecov-action@v2
       with:
         token: ${{ secrets.CODECOV_TOKEN }}
         file: ./coverage.xml
@@ -84,7 +84,7 @@ jobs:
       run: tox -e external
 
     - name: Coverage
-      uses: codecov/codecov-action@v1
+      uses: codecov/codecov-action@v2
       with:
         token: ${{ secrets.CODECOV_TOKEN }}
         file: ./coverage.xml
@@ -118,7 +118,7 @@ jobs:
       run: tox -e petab
 
     - name: Coverage
-      uses: codecov/codecov-action@v1
+      uses: codecov/codecov-action@v2
       with:
         token: ${{ secrets.CODECOV_TOKEN }}
         file: ./coverage.xml
@@ -152,7 +152,7 @@ jobs:
         run: tox -e mac
 
       - name: Coverage
-        uses: codecov/codecov-action@v1
+        uses: codecov/codecov-action@v2
         with:
           token: ${{ secrets.CODECOV_TOKEN }}
           file: ./coverage.xml
@@ -248,7 +248,7 @@ jobs:
       run: tox -e migrate
 
     - name: Coverage
-      uses: codecov/codecov-action@v1
+      uses: codecov/codecov-action@v2
       with:
         token: ${{ secrets.CODECOV_TOKEN }}
         file: ./coverage.xml

CHANGELOG.rst

@@ -8,6 +8,12 @@ Release Notes
 ...........
 
 
+0.12.3 (2022-04-05)
+-------------------
+
+* Document custom priors (#559)
+
+
 0.12.2 (2022-03-25)
 -------------------
 

README.rst

@@ -22,6 +22,7 @@ Written in Python with support for especially R and Julia.
 - **Contact:** https://pyabc.rtfd.io/en/latest/about.html
 - **Bug reports:** https://github.com/icb-dcm/pyabc/issues
 - **Source code:** https://github.com/icb-dcm/pyabc
+- **Cite:** https://pyabc.rtfd.io/en/latest/cite.html
 
 .. |CI| image:: https://github.com/ICB-DCM/pyABC/workflows/CI/badge.svg
    :target: https://github.com/ICB-DCM/pyABC/actions

doc/examples.rst

@@ -36,6 +36,7 @@ Algorithms and features
 
    examples/early_stopping.ipynb
    examples/resuming.ipynb
+   examples/custom_priors.ipynb
    examples/adaptive_distances.ipynb
    examples/informative.ipynb
    examples/aggregated_distances.ipynb

doc/examples/custom_priors.ipynb

@@ -3,33 +3,28 @@
 Install und Upgrade
 ===================
 
-
 Preparation
 -----------
 
 This package requires Python 3.8 or later.
-The package is tested on Linux (using Travis continuous integration).
-
-Not all of the package's functionality is available for Microsoft Windows.
-As some of the multi-core parallelizations rely on forking,
-these won't work on Windows.
-However, most other parts of the
-package should work on Windows
-as well.
+The package is continuously tested on Linux, and in parts on iOS,
+via GitHub Actions.
 
+While many parts of the package should work on Microsoft Windows
+as well, in particular the multi-core parallelizations rely on forking,
+which won't be available.
+Still, the main development focus is on Linux.
 
 My system's Python distribution is outdated, what now?
 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
 
-Several Python distributions can coexist on a single system.
+Several Python distributions can coexist on a system.
 If you don't have access to a recent Python version via your
-system's package manager (this might be the case for old
-Debian or Ubuntu operating systems),
-it is recommended to install the latest version of the
+system's package manager (may be the case for old systems),
+we recommend to install the latest version of the
 `Anaconda Python 3 distribution <https://www.continuum.io/downloads>`_.
 See also: :ref:`anacondaCluster`.
 
-
 PIP Installation
 ----------------
 
@@ -40,10 +35,8 @@ The package can be installed via pip.::
 
     pip install pyabc
 
-
 into your system's Python distribution. This requires usually root access.
 
-
 Install as user into your home directory (recommended)
 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
 
@@ -56,20 +49,15 @@ home directory with::
 
    pip install --user pyabc
 
-
-
 GIT Installation
 ----------------
 
 If you want the bleeding edge version, install directly from github::
 
    pip install git+https://github.com/icb-dcm/pyabc.git
 
-
-
 .. _upgrading:
 
-
 Upgrading
 ---------
 
@@ -78,61 +66,36 @@ pyABC version, use::
 
     pip install --upgrade pyabc
 
-
 instead of ``pip install``.
 You can also consult the `pip documentation <https://pip.pypa.io/en/stable/>`_
 on how to manage packages.
 If you installed pyABC into your
 home directory with
 ``pip install --user pyabc``, then upgrade also with the ``--user`` flag::
 
-
     pip install --upgrade --user pyabc
 
-
-
-
 .. _anacondaCluster:
 
 Installing Anaconda on a Cluster environment
 --------------------------------------------
 
-We're assuming you're on a Linux environment.
-Use the most recent Anaconda Python 3.x distribution.
-As of writing this documentation, this is the
-`Anaconda Python 3.6 <https://www.continuum.io/downloads>`_ distribution.
-To install it, run::
-
-   wget https://repo.continuum.io/archive/Anaconda3-4.4.0-Linux-x86_64.sh
-
-to download the installer. To execute the installer run::
-
-   bash Anaconda3-4.4.0-Linux-x86_64.sh
-
-and follow the guided installation process (i.e. approve the license
-and tell the installer where to install it to). You might want to replace
-the "4.4.0" by the most recent version of Anaconda.
-Find out on the `Anaconda Download <https://www.continuum.io/downloads>`_
-page which one it is.
-
-
-.. note::
+To install `Anaconda <https://anaconda.com/products/distribution>`_, run::
 
-    The Anaconda installer asks you at the end of the installation whether
-    you want to use Anaconda Python as your default Python:: bash
+   wget https://repo.anaconda.com/archive/Anaconda3-2021.11-Linux-x86_64.sh
+   bash Anaconda3-2021.11-Linux-x86_64.sh
 
-       Do you wish the installer to prepend the Anaconda3 install location
-       to PATH in your /home/username/.bashrc ? [yes|no]
-       [no] >>>
+and follow the installation guide.
+Replace the "2021.11" by the most recent version of Anaconda, see
+https://repo.anaconda.com/archive.
 
-    If you answer yes, the path to the Anaconda installation is prepended to
-    your ``PATH`` environment variable and subsequent calls to ``pip``
-    (see below) use the Anaconda Python pip (check with the command
-    ``which pip``).
-    If you answer no, you need to ensure manually, that the correct Python
-    installation is used.
-    Just saying "yes" here might safe you from some difficulties later on.
+`Miniconda <https://docs.conda.io/en/latest/miniconda.html>`_
+provides an alternative, minimal installer for conda, including
+only conda, Python, and some core and useful packages. Install the latest
+version via::
 
+  wget https://repo.anaconda.com/miniconda/Miniconda3-latest-Linux-x86_64.sh
+  bash Miniconda3-latest-Linux-x86_64.sh
 
 .. _install-optional:
 

doc/installation.rst

@@ -99,6 +99,7 @@
 from .random_variables import (
     RV,
     Distribution,
+    DistributionBase,
     LowerBoundDecorator,
     RVBase,
     RVDecorator,

pyabc/__init__.py

@@ -27,7 +27,7 @@ class RVBase(ABC):
         distributions are not pickleable.
         This class is really a very thin wrapper around ``scipy.stats``
         distributions to make them pickleable.
-        It is important to be able to pickle them to execute the ACBSMC
+        It is important to be able to pickle them to execute the ABCSMC
         algorithm in a distributed cluster
         environment
     """
@@ -192,7 +192,7 @@ class RVDecorator(RVBase):
 
     It stores the decorated random variable in ``self.component``
 
-    Overwrite the method ``decorator_repr`` the represent the decorator type.
+    Overwrite the method ``decorator_repr`` to represent the decorator type.
     The decorated variable will then be automatically included in
     the call to ``__repr__``.
 
@@ -310,16 +310,42 @@ def cdf(self, x, *args, **kwargs):
         return (self.component.cdf(x) - lower_mass) / (1 - lower_mass)
 
 
-class Distribution(ParameterStructure):
-    """Distribution of parameters for a model.
+class DistributionBase(ABC):
+    """Distribution of parameters for a model, abstract base class.
 
     A distribution is a collection of RVs and/or distributions.
+
+    This should be used to define a prior.
+    """
+
+    @abstractmethod
+    def rvs(self, *args, **kwargs) -> Parameter:
+        """Sample from joint distribution.
+
+        Returns
+        -------
+        parameter: Parameter
+            A parameter which was sampled.
+        """
+
+    @abstractmethod
+    def pdf(self, x: Union[Parameter, dict]):
+        """Get probability density at point `x`.
+
+        Parameters
+        ----------
+        x : Union[Parameter, dict]
+            Evaluate at the given Parameter ``x``.
+        """
+
+
+class Distribution(DistributionBase, ParameterStructure):
+    """Distribution of parameters for a model assuming independence.
+
     Essentially something like a dictionary
     of random variables or distributions.
     The variables from which the distribution is initialized are
     independent.
-
-    This should be used to define a prior.
     """
 
     def __repr__(self):
@@ -394,7 +420,6 @@ def rvs(self, *args, **kwargs) -> Parameter:
 
         Returns
         -------
-
         parameter: Parameter
             A parameter which was sampled.
         """

pyabc/random_variables.py

@@ -1 +1 @@
-__version__ = '0.12.2'
+__version__ = '0.12.3'

pyabc/version.py

@@ -23,6 +23,7 @@ nbs_2=(
   "data_plots"
   "discrete_parameters"
   "optimal_threshold"
+  "custom_priors"
 )
 
 # All notebooks