Re: DiscreteUniformDistribution PDF isn't piecewise?
- To: mathgroup at smc.vnet.net
- Subject: [mg107702] Re: DiscreteUniformDistribution PDF isn't piecewise?
- From: DrMajorBob <btreat1 at austin.rr.com>
- Date: Tue, 23 Feb 2010 08:05:06 -0500 (EST)
- References: <hlqu3v$go8$1@smc.vnet.net> <201002220807.DAA28826@smc.vnet.net>
- Reply-to: drmajorbob at yahoo.com
Fair enough. If WRI provides both the PDF and the CDF, we needn't derive
one from the other.
It's unfortunate if one tries to do so and it fails, but we've had this
discussion, so it's no longer a surprise. (For some it never was, I'm
sure.)
The lack of a Domain function isn't so terrible either, since, if you've
set dist = (some distribution) when I wasn't looking, I can discover the
distribution with Head and the domain with Part.
Bobby
On Mon, 22 Feb 2010 22:41:26 -0600, Peter Falloon <pfalloon at gmail.com>
wrote:
>>
>> The problem with a PMF is one we've already seen: 1/6& isn't a PMF (or
>> PDF)
>> at all, unless we know its accompanying domain. It could be uniform
>> discrete
>> on ANY 6 numbers or other objects.
>>
>
> I agree; what I'm trying to say is that a possible solution for a PMF (as
> opposed to a PDF) is as you suggested, except replacing DiracDelta by
> KroneckerDelta:
>
> myPDF[DiscreteUniformDistribution[{imin_Integer,imax_Integer}], x_]
> :=
>
> (1/(imax-imin+1))*Sum[KroneckerDelta[x,i], {i,imin,imax}]
>
> This has the desired behaviour for a PMF (which, according to the
> documention, is what the Mathematica implementation seeks to provide for
> discrete distributions such as this one) -- whereas the DiracDelta
> version
> would be appropriate for a PDF.
>
> Cheers,
> Peter.
--
DrMajorBob at yahoo.com
- References:
- Re: DiscreteUniformDistribution PDF isn't piecewise?
- From: "Nasser M. Abbasi" <nma@12000.org>
- Re: DiscreteUniformDistribution PDF isn't piecewise?