Merge 47eaf1a into 5360c08

.travis.yml

@@ -7,9 +7,9 @@ git:
   quiet: true
 
 python:
+  - 3.10
   - 3.9
   - 3.8
-  - 3.7
 
 install:
   - pip3 install --upgrade pip  # all three OSes agree about 'pip3'

doc/source/examples.rst

@@ -168,13 +168,15 @@ weight matrix by specifying the ``weight`` argument:
     ...                    betas=(0.1, 0.1), method='openmp', weights=weights)
     >>> weights2  # doctest: +ELLIPSIS
     <xarray.DataArray (outcomes: 8, cues: 15)>
-    array(...
+    array([[ 0.24...
     ...
-    ...)
+    ...]])
     Coordinates:
-      * outcomes  (outcomes) <U6 ...
-      * cues      (cues) <U2 ...
-    Attributes: ...
+      * outcomes  (outcomes) <U6 'hand' 'plural'...
+      * cues      (cues) <U2 '#h' 'ha' 'an' 'nd'...
+    Attributes:
+        date:...
+        event_path:...
     ...
 
 As you may have noticed already, :py:mod:`pyndl.ndl.ndl` provides you with meta

pyndl/__init__.py

@@ -25,7 +25,7 @@
               'David-Elias Künstle, Lennard Schneider, '
               'Elnaz Shafaei-Bajestan')
 __author_email__ = 'konstantin.sering@uni-tuebingen.de'
-__version__ = '0.8.1'
+__version__ = '0.8.2'
 __license__ = 'MIT'
 __description__ = ('Naive discriminative learning implements learning and '
                    'classification models based on the Rescorla-Wagner '
@@ -38,9 +38,9 @@
     'Operating System :: POSIX :: Linux',
     'Operating System :: MacOS',
     'Programming Language :: Python',
-    'Programming Language :: Python :: 3.7',
     'Programming Language :: Python :: 3.8',
     'Programming Language :: Python :: 3.9',
+    'Programming Language :: Python :: 3.10',
     'Programming Language :: Python :: 3 :: Only',
     'Topic :: Scientific/Engineering',
     'Topic :: Scientific/Engineering :: Artificial Intelligence',

pyndl/wh.py

@@ -2,7 +2,15 @@
 pyndl.wh
 --------
 
-*pyndl.wh* provides functions in order to train WH models
+*pyndl.wh* provides functions in order to train Widrow-Hoff (WH) models. In contrast
+to the Rescorla-Wagner (RW) models in the WH models can not only have binary
+cues and outcomes, but can code gradual intesities in there cues and outcomes
+as real, continues values.
+
+It is possible to calculate weights for continuous cues or continues outcomes,
+while keeping the outcomes respectively cues binary. Finally, it is possible to
+have both sides, cues and outcomes, to be continues and calculate the
+Widrow-Hoff learning rule between them.
 
 """
 from collections import defaultdict, OrderedDict
@@ -45,8 +53,17 @@ def wh(events, eta, *, cue_vectors=None, outcome_vectors=None,
         verbose=False, temporary_directory=None,
         events_per_temporary_file=10000000):
     """
-    Calculate the weights for all events using the Widrow-Hoff learning rule
-    and training as outcomes on sematic vectors in semantics.
+    Calculate the weights for all events using the Widrow-Hoff learning rule in
+    three different flavours.
+
+    In the first flavour, cues and outcomes both are vectors and the names in
+    the eventfiles refer to these vectors. The vectors for all cues and
+    outcomes are given as an xarrayDataArray with the arguments `cue_vectors`
+    and `outcome_vectors'.
+
+    In the second and third flavour, only the cues or only the outcomes are
+    treated as vectors but the ones not being treated as vectors are still
+    considered being present or not being present in a binary way.
 
     This is a parallel python implementation using numpy, multithreading and
     the binary format defined in preprocess.py.

tox.ini

@@ -1,5 +1,5 @@
 [tox]
-envlist = py{37,38,39}-test, checkstyle, documentation
+envlist = py{38,39,310}-test, checkstyle, documentation
 
 [testenv]
 usedevelop = True