commons-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Phil Steitz <>
Subject Re: [math] Stochastic Process
Date Tue, 04 Jun 2013 14:18:59 GMT
On 6/3/13 12:06 PM, Thomas Neidhart wrote:
> On 06/03/2013 04:10 PM, Phil Steitz wrote:
>> On 6/3/13 4:18 AM, Thomas Neidhart wrote:
>>> Hi,
>>> to start working on the Monte Carlo engine (see MATH-463) I would like to
>>> break this thing up in multiple pieces. One thing that could be added
>>> independently is the concept of a stochastic process (e.g. Wiener,
>>> BrownianMotion, ...).
>> +1
>>> The code in the contribution is already a pretty good start, but the
>>> question would be where to put it. We do not yet have a stochastic base
>>> package, and random is also not such a good fit imho.
>>> I see various options:
>>>  - random.process: well it models a random process ...
>> random was originally intended to just house random data generation
>> stuff.  I would see stochastic processes as logical clients of the
>> generators in random, but I see the logic here.
>>>  - stochastic.process: downside of adding another top-level package
>> It think this is probably the best.  I tend to favor shallow and
>> wide over deep and narrow because it makes it easier to find things
>> and leads to less head-scratching about why things are where they
>> are.  What are the other stochastic.x that you have in mind?
> I would also prefer this, but currently I do not have a clear idea of
> other packages that would reside beneath stochastic.

I don't either, which is why I asked.  What comes to mind
immediately is .simulation; but I am not sure you actually end up
with that (different sorts will go with the actual techniques used)
or .optimization (but that stuff is already in the optimization

>>>  - stat.process: well, statistics is a sub-group of stochastic so it would
>>> not be perfect
>> He he.  A probabilist, I assume :)  What you probably really mean is
>> probability.stochastic.process, probability.stat.*.  Too deep for me :)
> yes, I would not propose something like that ;-).
> Thomas
> ---------------------------------------------------------------------
> To unsubscribe, e-mail:
> For additional commands, e-mail:

To unsubscribe, e-mail:
For additional commands, e-mail:

View raw message