Merge pull request #129 from elfi-dev/install_docs

Updated installation instructions and other details in the docs

CONTRIBUTING.rst

@@ -15,7 +15,7 @@ Types of Contributions
 Report Bugs
 ~~~~~~~~~~~
 
-Report bugs at https://github.com/HIIT/elfi/issues.
+Report bugs at https://github.com/elfi-dev/elfi/issues.
 
 If you are reporting a bug, please include:
 
@@ -45,7 +45,7 @@ articles, and such.
 Submit Feedback
 ~~~~~~~~~~~~~~~
 
-The best way to send feedback is to file an issue at https://github.com/HIIT/elfi/issues.
+The best way to send feedback is to file an issue at https://github.com/elfi-dev/elfi/issues.
 
 If you are proposing a feature:
 
@@ -119,7 +119,7 @@ Before you submit a pull request, check that it meets these guidelines:
    your new functionality into a function with a docstring, and add the
    feature to the list in README.rst.
 3. The pull request should work for Python 2.7,  3.5 and later. Check
-   https://travis-ci.org/HIIT/elfi/pull_requests
+   https://travis-ci.org/elfi-dev/elfi/pull_requests
    and make sure that the tests pass for all supported Python versions.
 
 Tips

README.md

@@ -9,9 +9,9 @@ ELFI - Engine for Likelihood-Free Inference
 <img src="https://cloud.githubusercontent.com/assets/1233418/20178983/6e22ee44-a75c-11e6-8345-5934b55b9dc6.png" width="15%" align="right"></img>
 
 ELFI is a statistical software package written in Python for performing inference with 
-generative models. The term "likelihood-lree inference" refers to a family of inference
+generative models. The term "likelihood-free inference" refers to a family of inference
 methods that replace the use of the likelihood function with a data generating simulator 
-function. This is useful when the likelihood function is not computable or otherwise 
+function. This is useful when the likelihood function is not computable or otherwise
 available but it is possible to make simulations of the process.
 
 Other names or related approaches to likelihood-free inference include Approximative 
@@ -50,7 +50,6 @@ installation.
 
 ### Optional dependencies
 
-- `ipyparallel` for parallelization
 - `graphviz` for drawing graphical models ([Graphviz](http://www.graphviz.org))
 
 

docs/installation.rst

@@ -21,8 +21,14 @@ ELFI is currently tested only with Python 3.5. If you are new to Python, perhaps
 
 .. _Anaconda: https://www.continuum.io/downloads
 
-Currently it is required to use Distributed 1.14.3.
+Optional dependencies
+---------------------
 
+Optionally you may wish to install also the following packages:
+
+* `graphviz` for drawing graphical models (Graphviz_ must be installed separately)
+
+.. _Graphviz: http://www.graphviz.org
 
 Virtual environment using Anaconda
 ----------------------------------
@@ -55,19 +61,24 @@ You can either clone the public repository:
 
 .. code-block:: console
 
-    git clone https://github.com/HIIT/elfi.git
+    git clone https://github.com/elfi-dev/elfi.git
 
 Or download the `tarball`_:
 
 .. code-block:: console
 
-    curl  -OL https://github.com/HIIT/elfi/tarball/master
+    curl  -OL https://github.com/elfi-dev/elfi/tarball/master
+
+Note that for development it is recommended to base your work on the `dev` branch instead of `master`.
 
 Once you have a copy of the source, you can install it with:
 
 .. code-block:: console
 
-    python setup.py install
+   pip install -e .
+
+This will install ELFI along with its default requirements.
+
+.. _Github repo: https://github.com/elfi-dev/elfi
+.. _tarball: https://github.com/elfi-dev/elfi/tarball/master
 
-.. _Github repo: https://github.com/HIIT/elfi
-.. _tarball: https://github.com/HIIT/elfi/tarball/master

docs/readme.rst

@@ -5,9 +5,7 @@ ELFI is a statistical software package written in Python for Approximative Bayes
 
 .. _ABC: https://en.wikipedia.org/wiki/Approximate_Bayesian_computation
 
-The probabilistic inference model is defined as a directed acyclic graph, which allows for an intuitive means to describe inherent dependencies in the model. The inference pipeline is automatically parallelized with Dask_, which scales well from a desktop up to a cluster environment. The package includes functionality for input/output operations and visualization.
-
-.. _Dask: https://dask.pydata.org
+The probabilistic inference model is defined as a directed acyclic graph, which allows for an intuitive means to describe inherent dependencies in the model. The inference pipeline is automatically parallelized from multiple cores up to a cluster environment. ELFI also handles seeding the random number generators and storing of the generated data for you so that you can easily repeat or fine tune your inference. Additionally, the package includes functionality for visualization.
 
 Currently implemented ABC methods:
 
@@ -17,12 +15,12 @@ Currently implemented ABC methods:
 
 .. _BOLFI: http://jmlr.csail.mit.edu/papers/v17/15-017.html
 
-GitHub page: https://github.com/HIIT/elfi
+GitHub page: https://github.com/elfi-dev/elfi
 
 See examples under the notebooks_ directory to get started. Limited user-support may be asked from elfi-support.at.hiit.fi, but the `Gitter chat`_ is preferable.
 
-.. _notebooks: https://github.com/HIIT/elfi/tree/master/notebooks
-.. _Gitter chat: https://gitter.im/HIIT/elfi?utm_source=share-link&utm_medium=link&utm_campaign=share-link
+.. _notebooks: https://github.com/elfi-dev/notebooks
+.. _Gitter chat: https://gitter.im/elfi-dev/elfi?utm_source=share-link&utm_medium=link&utm_campaign=share-link
 
 Licenses:
 

docs/usage.rst

@@ -6,6 +6,7 @@ To use ELFI in a project::
 
     import elfi
 
-For tutorials, please see the Jupyter Notebooks under the `notebooks directory`_.
+For tutorials, please see the Jupyter Notebooks under the `notebooks directory`_. Feel free to add your own in the zoo_.
 
-.. _notebooks directory: https://github.com/HIIT/elfi/tree/master/notebooks
+.. _notebooks directory: https://github.com/elfi-dev/notebooks
+.. _zoo: https://github.com/elfi-dev/zoo

examples/ma2.py

@@ -24,12 +24,32 @@ def MA2(t1, t2, n_obs=100, batch_size=1, random_state=None):
 def autocov(x, lag=1):
     """Autocovariance assuming a (weak) univariate stationary process with mean 0.
     Realizations are in rows.
+
+    Parameters
+    ----------
+    x : np.array of size (n, m)
+    lag : int, optional
+
+    Returns
+    -------
+    C : np.array of size (n,)
     """
     C = np.mean(x[:, lag:]*x[:, :-lag], axis=1)
     return C
 
 
 def discrepancy(x, y):
+    """Euclidean distance between data.
+
+    Parameters
+    ----------
+    x : tuple of np.arrays of size (n,)
+    y : tuple of np.arrays of size (n,)
+
+    Returns
+    -------
+    d : np.array of size (n,)
+    """
     d = np.linalg.norm(np.column_stack(x) - np.column_stack(y), ord=2, axis=1)
     return d
 

requirements-dev.txt

@@ -1,16 +1,16 @@
 # Testing
-pytest==3.0.3
-tox==2.4.1
-coverage==4.2
-pytest-cov==2.4.0
+pytest>=3.0.3
+tox>=2.4.1
+coverage>=4.2
+pytest-cov>=2.4.0
 
 # Linting
-flake8==3.0.4
-pep8-naming==0.4.1
-flake8-docstrings==1.0.2
-isort==4.2.5
-flake8-isort==2.0.1
+flake8>=3.0.4
+pep8-naming>=0.4.1
+flake8-docstrings>=1.0.2
+isort>=4.2.5
+flake8-isort>=2.0.1
 
 # Documentation
-Sphinx==1.4.8
-watchdog==0.8.3
+Sphinx>=1.4.8
+watchdog>=0.8.3

requirements.txt

@@ -0,0 +1,6 @@
+numpy>=1.8
+scipy>=0.16.1
+matplotlib>=1.1
+GPy>=1.0.9
+networkX>=1.11
+ipyparallel>=6

setup.py

@@ -8,14 +8,8 @@
 
 packages = ['elfi'] + ['elfi.' + p for p in find_packages('elfi')]
 
-requirements = [
-                'numpy>=1.8',
-                'scipy>=0.16.1',
-                'matplotlib>=1.1',
-                'GPy>=1.0.9',
-                'networkX>=1.11',
-                'ipyparallel>=6.0'
-                ]
+with open('requirements.txt', 'r') as f:
+    requirements = f.read().splitlines()
 
 optionals = {
     'doc': ['Sphinx'],
@@ -30,7 +24,7 @@
     keywords='abc likelihood-free statistics',
     packages=packages,
     version=__version__,
-    author='HIIT',
+    author='ELFI authors',
     author_email='elfi-support@hiit.fi',
     url='http://elfi.readthedocs.io',