On Fri, Jun 23, 2017 at 2:58 AM, Alvaro Herrera
<alvherre@2ndquadrant.com> wrote:
> Ashutosh Bapat wrote:
>> On Mon, Jun 5, 2017 at 8:22 AM, atorikoshi
>> <torikoshi_atsushi_z2@lab.ntt.co.jp> wrote:
>> > Hi,
>> >
>> > I found below formula to compute selectivities, but
>> > I think the last Probability 'P(b=?)' should be 'P(c=?)'.
>> >
>> >> P(a=?,b=?,c=?) = P(a=?,b=?) * (d + (1-d)*P(b=?))
>> >
>> >
>> > Attached patch fixes it, and it also adds some spaces
>> > following another formula which is on line 86 and
>> > computes P(a=?, b=?).
>>
>> Agree. Also using "d" for "degree of functional dependence (b=>c) as
>> well is confusing. We are already using "d" for "degree of functional
>> dependence (a=>b). Here' patch to use "d'" instead of "d".
>
> Since the surrounding text uses single quotes to talk about each letter,
> I thought it was better to use a new letter (e) so that we don't require
> the "prime" notation, which would end up being either inconsistent,
> confusing, stupid-looking, or combinations thereof.
Makes sense.
>
> Also, your proposed text had a slight mistake: it's not (b=>c) that
> d' is for, but (a,b=>c).
Sorry for that.
>
> Pushed with those corrections.
Thanks.
--
Best Wishes,
Ashutosh Bapat
EnterpriseDB Corporation
The Postgres Database Company