| From: | "Tomas Vondra" <tv(at)fuzzy(dot)cz> |
|---|---|
| To: | "Gavin Flower" <GavinFlower(at)archidevsys(dot)co(dot)nz> |
| Cc: | "Tomas Vondra" <tv(at)fuzzy(dot)cz>, "Katharina Büchse" <katharina(dot)buechse(at)uni-jena(dot)de>, pgsql-hackers(at)postgresql(dot)org |
| Subject: | Re: two dimensional statistics in Postgres |
| Date: | 2014-11-06 11:17:43 |
| Message-ID: | 04ef58463197bcc46e78672c7a317bf1.squirrel@2.emaily.eu |
| Views: | Raw Message | Whole Thread | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-hackers |
Dne 6 Listopad 2014, 12:05, Gavin Flower napsal(a):
> On 06/11/14 23:57, Tomas Vondra wrote:
>> Dne 6 Listopad 2014, 11:50, Gavin Flower napsal(a):
>>> Could you store a 2 dimensional histogram in a one dimensional array:
>>> A[z] = value, where z = col * rowSize + row (zero starting index)?
>> How would that work for columns with different data types?
>>
>> Tomas
>>
> I implicitly assumed that all cells were the same size & type. However,
> I could devise a scheme to cope with columns of different types, given
> the relevant definitions - but this would obviously be more complicated.
>
> Also this method can be extended into higher dimensions.
Which is what I did in the "WIP: multivariate stats" ;-)
Tomas
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Heikki Linnakangas | 2014-11-06 11:23:28 | Re: WAL format and API changes (9.5) |
| Previous Message | Magnus Hagander | 2014-11-06 11:16:44 | Re: Order of views in stats docs |