[
https://issues.apache.org/jira/browse/MATH-1267?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14737533#comment-14737533
]
Gilles commented on MATH-1267:
------------------------------
{quote}
The design idea that I'm advocating comes from MVC theory, hence the references
to things you can do. I think it's nice that you can do these things and that
the design easily supports it, even though the classes are purely used for
computation ATM.
{quote}
What are the things that the design does not support?
{quote}
Even though there may not be a Neuron at every point in a square mesh /
coordinate system, we can still think of a Neuron as belonging to a row, column
coordinate, right?
{quote}
It depends.
If the network is static and has uniform density, yes. If not, no.
Even in the former case, in an hexagonal layout, the (row, column) concept is
not that useful I think (because the network is not the grid anymore).
Then when the network is adaptive:
Before:
{noformat}
*-----*-----X
| | |
*-----*-----*
{noformat}
Neuron "X" is at (0, 2).
After adding a neuron:
{noformat}
*-----*--*--X
| | |
*-----*-----*
{noformat}
Neuron "X" is at (0, 3).
Hence the "Neuron2D" coordinates of "X" have become wrong.
> Grid coordinate of "Neuron" that belongs to a "NeuronSquareMesh2D"
> ------------------------------------------------------------------
>
> Key: MATH-1267
> URL: https://issues.apache.org/jira/browse/MATH-1267
> Project: Commons Math
> Issue Type: Wish
> Reporter: Gilles
> Assignee: Gilles
> Priority: Minor
> Fix For: 4.0, 3.6
>
> Attachments: LocationFinder.java
>
>
> When the network layout is a 2D grid, it useful to be able to retrieve the
> grid coordinate (row, column) of a a neuron.
> I propose to define a new class "LocationFinder" (in package
> "o.a.c.m.ml.neuralnet.twod.util") that will provide the functionality through
> a "getLocation(Neuron n)" method.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
