Hi Aldo,
The documentation is pretty good. You can change the number of clusters
like so
my $k = $pdl->kmeans( {NCLUS => 10} );
There are also a few blog posts that show some of the process from a
learners' perspective
https://blogs.perl.org/users/enkidu/2019/12/its-beginning-to-look-a-lot-like-k-means.html
There's not a lot of examples out there, so if you can blog your experience
(good or bad), that'll add to the available docs for the next person.
dev.to gets a lot of exposure these days.
best of luck,
Boyd
On Wed, 22 Nov 2023 at 10:48, Aldobranti <[email protected]> wrote:
> my problem area is to analyse electrical power switching events and tie
> back readings of reactive and resistive loads aiming to point back at
> specific items of plant hardware.
>
> So I thought cluster analysis and got a steer towards PDL
>
> In a perl script I write
>
> use PDL::Stats::KMeans;
>
> ...my $sql = "select resistv_chg,reactv_chg from metered_events where
> matched_event_id is not null";my $pdl = rdbi2D( $dbh, $sql ); # pdl info
> tells me that i have a 2D array of ~ 1400 x 2my $k = $pdl->kmeans( {} );
>
> First up I see that this is not unrestrained -- the default of NCLUS is 3
> and [seems to me] that I am being defeated in trying to identify between 10
> and 20 clusters associated with the electrical characteristics of the 10 -
> 20 items in my plant
>
> being a simple minded chap all I want is an array[10] of mean and stdvars
> for the data set
>
> Could anybody spare the time to get me started?
> Aldo
> --
> Aldobranti: A performance in digital social media and large sheets of
> photographic film https://aldobranti.org
>
>
>
> _______________________________________________
> pdl-general mailing list
> [email protected]
> https://lists.sourceforge.net/lists/listinfo/pdl-general
>
--
Boyd Duffee
No pain, no French bread - Teresa Monachino
_______________________________________________
pdl-general mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/pdl-general
