Release 0.7 (#252)

* Fix undefined variable

* Bump version to 0.7, update API docs topics

* Add scipy.linalg to mocked libraries

* Update API docs

Author: vuolleko

CHANGELOG.rst

@@ -1,15 +1,16 @@
 Changelog
 =========
 
-dev
----
-- Added new example: the stochastic Lotka-Volterra model
-- Fix methods.bo.utils.minimize to be strictly within bounds
-- Implemented the Two Stage Procedure, a method of summary-statistics diagnostics  
+0.7 (2017-11-30)
+----------------
+
 - Added the MaxVar acquisition method
 - Added the RandMaxVar acquisition method
-- Fix elfi.Distance to support scipy 1.0.0
 - Added the ExpIntVar acquisition method
+- Implemented the Two Stage Procedure, a method of summary-statistics diagnostics
+- Added new example: the stochastic Lotka-Volterra model
+- Fix methods.bo.utils.minimize to be strictly within bounds
+- Fix elfi.Distance to support scipy 1.0.0
 
 0.6.3 (2017-09-28)
 ------------------

README.md

@@ -1,4 +1,4 @@
-**Version 0.6.3 released!** See the CHANGELOG and [notebooks](https://github.com/elfi-dev/notebooks).
+**Version 0.7 released!** See the CHANGELOG and [notebooks](https://github.com/elfi-dev/notebooks).
 
 **NOTE:** For the time being NetworkX 2 is incompatible with ELFI.
 

docs/api.rst

@@ -7,6 +7,8 @@ Modelling API
 -------------
 Below is the API for creating generative models.
 
+.. currentmodule:: .
+
 .. autosummary::
    elfi.ElfiModel
 
@@ -28,16 +30,12 @@ Below is the API for creating generative models.
 
 **Other**
 
-.. currentmodule:: elfi.model.elfi_model
-
 .. autosummary::
    elfi.new_model
    elfi.load_model
    elfi.get_default_model
    elfi.set_default_model
 
-.. currentmodule:: elfi.visualization.visualization
-
 .. autosummary::
    elfi.draw
 
@@ -52,6 +50,7 @@ Below is a list of inference methods included in ELFI.
    elfi.BayesianOptimization
    elfi.BOLFI
 
+
 **Result objects**
 
 .. currentmodule:: elfi.methods.results
@@ -62,9 +61,10 @@ Below is a list of inference methods included in ELFI.
    SmcSample
    BolfiSample
 
+
 **Post-processing**
 
-.. currentmodule:: elfi
+.. currentmodule:: .
 
 .. autosummary::
    elfi.adjust_posterior
@@ -74,20 +74,40 @@ Below is a list of inference methods included in ELFI.
 .. autosummary::
    LinearAdjustment
 
+
+**Diagnostics**
+
+.. currentmodule:: .
+
+.. autosummary::
+   elfi.TwoStageSelection
+
+
+**Acquisition methods**
+
+.. currentmodule:: elfi.methods.bo.acquisition
+
+.. autosummary::
+   LCBSC
+   MaxVar
+   RandMaxVar
+   ExpIntVar
+   UniformAcquisition
+
 Other
 -----
 
 **Data pools**
 
+.. currentmodule:: .
+
 .. autosummary::
    elfi.OutputPool
    elfi.ArrayPool
 
 
 **Module functions**
 
-.. currentmodule:: elfi
-
 .. autosummary::
    elfi.get_client
    elfi.set_client
@@ -102,6 +122,7 @@ Other
    elfi.tools.external_operation
 
 
+
 Class documentations
 --------------------
 
@@ -146,17 +167,15 @@ Modelling API classes
 
 **Other**
 
-.. currentmodule:: elfi.model.elfi_model
+.. autofunction:: elfi.new_model
 
-.. automethod:: elfi.new_model
+.. autofunction:: elfi.load_model
 
-.. automethod:: elfi.get_current_model
+.. autofunction:: elfi.get_default_model
 
-.. automethod:: elfi.set_current_model
+.. autofunction:: elfi.set_default_model
 
-.. currentmodule:: elfi.visualization.visualization
-
-.. automethod:: elfi.visualization.visualization.nx_draw
+.. autofunction:: elfi.draw
 
 .. This would show undocumented members :undoc-members:
 
@@ -180,11 +199,11 @@ Inference API classes
    :members:
    :inherited-members:
 
-.. currentmodule:: elfi.methods.results
-
 
 **Result objects**
 
+.. currentmodule:: elfi.methods.results
+
 .. autoclass:: OptimizationResult
    :members:
    :inherited-members:
@@ -204,22 +223,55 @@ Inference API classes
 
 **Post-processing**
 
-.. currentmodule:: elfi
+.. currentmodule:: .
 
-.. automethod:: elfi.adjust_posterior
+.. autofunction:: elfi.adjust_posterior
 
 .. currentmodule:: elfi.methods.post_processing
 
 .. autoclass:: LinearAdjustment
    :members:
    :inherited-members:
 
+**Diagnostics**
+
+.. currentmodule:: .
+
+.. autoclass:: elfi.TwoStageSelection
+   :members:
+   :inherited-members:
+
+**Acquisition methods**
+
+.. currentmodule:: elfi.methods.bo.acquisition
+
+.. autoclass:: LCBSC
+   :members:
+   :inherited-members:
+
+.. autoclass:: MaxVar
+   :members:
+   :inherited-members:
+
+.. autoclass:: RandMaxVar
+   :members:
+   :inherited-members:
+
+.. autoclass:: ExpIntVar
+   :members:
+   :inherited-members:
+
+.. autoclass:: UniformAcquisition
+   :members:
+   :inherited-members:
 
 Other
 .....
 
 **Data pools**
 
+.. currentmodule:: .
+
 .. autoclass:: elfi.OutputPool
    :members:
    :inherited-members:
@@ -231,11 +283,9 @@ Other
 
 **Module functions**
 
-.. currentmodule:: elfi
-
-.. automethod:: elfi.get_client
+.. autofunction:: elfi.get_client
 
-.. automethod:: elfi.set_client
+.. autofunction:: elfi.set_client
 
 
 **Tools**

docs/conf.py

@@ -30,9 +30,9 @@ def __getattr__(cls, name):
     MOCK_MODULES = [
         'pygtk', 'gtk', 'gobject', 'argparse', 'numpy', 'pandas', 'scipy', 'unqlite', 'dask',
         'distributed', 'distributed.client', 'graphviz', 'matplotlib', 'sobol_seq', 'GPy',
-        'dask.delayed', 'scipy.optimize', 'scipy.stats', 'scipy.spatial', 'scipy.sparse',
-        'matplotlib.pyplot', 'numpy.random', 'networkx', 'ipyparallel', 'numpy.lib',
-        'numpy.lib.format', 'sklearn.linear_model'
+        'dask.delayed', 'scipy.linalg', 'scipy.optimize', 'scipy.stats', 'scipy.spatial',
+        'scipy.sparse', 'scipy.special', 'matplotlib.pyplot', 'numpy.random', 'networkx',
+        'ipyparallel', 'numpy.lib', 'numpy.lib.format', 'sklearn.linear_model'
     ]
     sys.modules.update((mod_name, Mock()) for mod_name in MOCK_MODULES)
 

docs/usage/external.rst

@@ -26,7 +26,8 @@ Let's begin by importing some libraries that we will be using:
     import scipy.stats as ss
     
     import elfi
-    import elfi.examples
+    import elfi.examples.bdm
+    import elfi.examples.ma2
     
     %matplotlib inline
 

docs/usage/parallelization.rst

@@ -252,4 +252,3 @@ Remember to stop the ipcluster when done
 
     2017-07-19 16:20:58.662 [IPClusterStop] Stopping cluster [pid=21020] with [signal=<Signals.SIGINT: 2>]
 
-

elfi/__init__.py

@@ -5,14 +5,13 @@
 package for likelihood-free inference (LFI) such as Approximate Bayesian
 Computation (ABC).
 """
-import logging
-logging.basicConfig(level=logging.INFO)
 
 import elfi.clients.native
 
 import elfi.methods.mcmc
 import elfi.model.tools as tools
 from elfi.client import get_client, set_client
+from elfi.methods.diagnostics import TwoStageSelection
 from elfi.methods.parameter_inference import *
 from elfi.methods.post_processing import adjust_posterior
 from elfi.model.elfi_model import *
@@ -25,4 +24,4 @@
 __email__ = 'elfi-support@hiit.fi'
 
 # make sure __version_ is on the last non-empty line (read by setup.py)
-__version__ = '0.6.3'
+__version__ = '0.7'

elfi/methods/diagnostics.py

@@ -19,14 +19,15 @@ class TwoStageSelection:
     or provide some already combined summary statistics as prepared_ss.
 
     The rationale of the Two Stage procedure procedure is the following:
-    - First, the module computes or accepts the combinations of the candidate summary statistics;
+
+    - First, the module computes or accepts the combinations of the candidate summary statistics.
     - In Stage 1, each summary-statistics combination is evaluated using the
-    Minimum Entropy algorithm;
+      Minimum Entropy algorithm.
     - In Stage 2, the minimum-entropy combination is selected,
-    and the `closest' datasets are identified;
+      and the 'closest' datasets are identified.
     - Further in Stage 2, for each summary-statistics combination,
-    the mean root sum of squared errors (MRSSE) is calculated over all `closest datasets',
-    and the minimum-MRSSE combination is chosen as the one with the optimal performance.
+      the mean root sum of squared errors (MRSSE) is calculated over all 'closest datasets',
+      and the minimum-MRSSE combination is chosen as the one with the optimal performance.
 
     References
     ----------
@@ -109,7 +110,7 @@ def run(self, n_sim, n_acc=None, n_closest=None, batch_size=1, k=4):
         n_acc : int, optional
             Number of the accepted ABC-rejection simulations.
         n_closest : int, optional
-            Number of the `closest' datasets
+            Number of the 'closest' datasets
             (i.e., the closest n simulation datasets w.r.t the observations).
         batch_size : int, optional
             Number of samples per batch.
@@ -136,6 +137,7 @@ def run(self, n_sim, n_acc=None, n_closest=None, batch_size=1, k=4):
         # preserve the parameters (thetas) corresponding to the `closest' datasets.
         thetas = {}
         E_me = np.inf
+        names_ss_me = []
         for set_ss in self.ss_candidates:
             names_ss = [ss.__name__ for ss in set_ss]
             thetas_ss = self._obtain_accepted_thetas(set_ss, n_sim, n_acc, batch_size)
@@ -151,8 +153,9 @@ def run(self, n_sim, n_acc=None, n_closest=None, batch_size=1, k=4):
         logger.info('\nThe minimum entropy of %f was found in %s.\n' % (E_me, names_ss_me))
 
         # Find the summary-statistics combination with
-        # the minimum mean root sum of squared error (MRRSE).
+        # the minimum mean root sum of squared error (MRSSE).
         MRSSE_min = np.inf
+        names_ss_MRSSE = []
         for set_ss in self.ss_candidates:
             names_ss = [ss.__name__ for ss in set_ss]
             MRSSE_ss = self._calc_MRSSE(set_ss, thetas_closest, thetas[set_ss])