Re: STATISTICS

Hi.

Thanks for your message about the problem with discrete random
variables.

| My idea was to allow all <ci> with attribute type=discrete_random_variable to 
| have a <sep/> such that we are first given x_i and then f(x_i). But is this 
| MathML??
| 
| Something like this:
| 
| <apply><mean/>
|   <cn type="discrete_random_variable"> 1<sep/> .2 </cn>
|   <cn type="discrete_random_variable"> 4 <sep/> .3 </cn>
|   <cn type="discrete_random_variable"> 6 <sep/> .5 </cn>
|  </apply>

The short answer to your question is "No, this isn't official
MathML".  In the 1.0 spec, there are not very many official "types"
defined. However, my thinking is that you shouldn't let this get in
your way.

I agree that there is a problem at the moment, and I think you have a
reasonable solution.  He always knew that problems like this would
come up as people actually began implementing software.  So in the
short term, I would suggest going ahead and implementing as you
suggest, and then lobbying to get it officially added.

The kickoff meeting for the rechartered W3C Math working group is next
week, and one of the main agenda items is to work out a procedure for
making corrections and additions to the spec, just exactly because of
problems like the one you identify.  Someone from the group will be
making a posting to this list in a couple of weeks with more news.
At any rate, we will talk about your proposal.

--Robert Miner

----------------------------------------------------------------
Robert Miner                          http://www.webeq.com
Geometry Technologies, Inc.           email: rminer@geomtech.com
                                      phone: 651-223-2884
----------------------------------------------------------------

Received on Thursday, 15 October 1998 16:15:09 UTC