__init__.py :  » Math » Modular-toolkit-for-Data-Processing » MDP-2.6 » mdp » hinet » Python Open Source

Home
Python Open Source
1.3.1.2 Python
2.Ajax
3.Aspect Oriented
4.Blog
5.Build
6.Business Application
7.Chart Report
8.Content Management Systems
9.Cryptographic
10.Database
11.Development
12.Editor
13.Email
14.ERP
15.Game 2D 3D
16.GIS
17.GUI
18.IDE
19.Installer
20.IRC
21.Issue Tracker
22.Language Interface
23.Log
24.Math
25.Media Sound Audio
26.Mobile
27.Network
28.Parser
29.PDF
30.Project Management
31.RSS
32.Search
33.Security
34.Template Engines
35.Test
36.UML
37.USB Serial
38.Web Frameworks
39.Web Server
40.Web Services
41.Web Unit
42.Wiki
43.Windows
44.XML
Python Open Source » Math » Modular toolkit for Data Processing 
Modular toolkit for Data Processing » MDP 2.6 » mdp » hinet » __init__.py
"""Hierarchical Networks Package.

This package makes it possible to construct graph-like Node structures,
especially hierarchical networks.

The most important building block is the new Layer node, which works as an 
horizontal version of flow. It encapsulates a list of Nodes, which are trained
and executed in parallel. 
For example we can take two Nodes with 100 dimensional input to
construct a layer with a 200 dimensional input. The first half of the input
data is automatically fed into the first Node, the second half into the second
Node.

Since one might also want to use Flows (i.e. vertical stacks of Nodes) in a
Layer, a wrapper class for Nodes is provided.
The FlowNode class wraps any Flow into a Node, which can then be used like any 
other Node. Together with the Layer this allows you to combine Nodes both
horizontally and vertically. Thereby one can in principle realize
any feed-forward network topology.

For a hierarchical networks one might want to route the different parts of the
data to different Nodes in a Layer in complicated ways. This is done by a
Switchboard that handles all the routing.
Defining the routing manually can be quite tedious, so one can derive subclasses
for special routing situations. One such subclass for 2d image data is provided. 
It maps the data according to rectangular overlapping 2d input areas. One can 
then feed the output into a Layer and each Node will get the correct input.
"""

from flownode import FlowNode
from layer import Layer,SameInputLayer,CloneLayer
from switchboard import (
    Switchboard, SwitchboardException, MeanInverseSwitchboard,
    ChannelSwitchboard,
    Rectangular2dSwitchboard, Rectangular2dSwitchboardException,
    DoubleRect2dSwitchboard, DoubleRect2dSwitchboardException,
    DoubleRhomb2dSwitchboard, DoubleRhomb2dSwitchboardException
)
from hinet_translator import (
    HiNetTranslator, HiNetHTMLTranslator, HiNetXHTMLTranslator,
    HINET_STYLE, NewlineWriteFile, show_flow
)
from switchboard_factory import (
    get_2d_image_switchboard, FactoryExtensionChannelSwitchboard,
    FactoryRectangular2dSwitchboard, FactoryDoubleRect2dSwitchboard,
    FactoryDoubleRhomb2dSwitchboard
)

del flownode
del layer
del switchboard
del hinet_translator
del switchboard_factory

__all__ = ['FlowNode', 'Layer', 'SameInputLayer', 'CloneLayer',
           'Switchboard', 'SwitchboardException', 'ChannelSwitchboard'
           'Rectangular2dSwitchboard', 'Rectangular2dSwitchboardException',
           'DoubleRect2dSwitchboard', 'DoubleRect2dSwitchboardException',
           'DoubleRhomb2dSwitchboard', 'DoubleRhomb2dSwitchboardException',
           'HiNetTranslator', 'HiNetHTMLTranslator', 'HiNetXHTMLTranslator',
           'HINET_STYLE', 'NewlineWriteFile', 'show_flow',
           'get_2d_image_switchboard'
           ]
www.java2java.com | Contact Us
Copyright 2009 - 12 Demo Source and Support. All rights reserved.
All other trademarks are property of their respective owners.