On 5 July 2011 17:32, Andrea Aime wrote:
> Or are you already the one? :-)
Me ???!!! I think not :-)
Michael
--
All of the data generated in your IT infrastructure is seriously valuable.
Why? It contains a definitive r
On Tue, Jul 5, 2011 at 9:14 AM, Michael Bedward
wrote:
> Hi Andrea,
>
> > On the fly compilation of classes will put the results in the permanent
> > generation,
> > depending on the garbage collector chosen it will result in a memory leak
> in
> > the permanent generation (once put there there mi
Hi Andrea,
> On the fly compilation of classes will put the results in the permanent
> generation,
> depending on the garbage collector chosen it will result in a memory leak in
> the permanent generation (once put there there might be no way to reclaim
> them).
> This actually depends on what hap
On Tue, Jul 5, 2011 at 5:52 AM, Michael Bedward
wrote:
> Hi again Andrea,
>
> Jody and I chatted on IRC a while ago about how using these functions
> with rasters and how bad this would be :) We were talking in
> particular about InterpolateFunction and wondered about compiling the
> interpolatio
When he says we it is more that he hurt my mind talking about byte code and
I ran and hid in a corner. I actually think it is really and interesting
idea; but would love to see some benchmarking to justify the complexity /
effort / magic.
Jody
On Tue, Jul 5, 2011 at 1:52 PM, Michael Bedward
wrote
Hi again Andrea,
Jody and I chatted on IRC a while ago about how using these functions
with rasters and how bad this would be :) We were talking in
particular about InterpolateFunction and wondered about compiling the
interpolation expression to bytecode using Janino. What do you think ?
Michael
On 4 July 2011 20:26, Andrea Aime wrote:
> I looked into the SE spec for RasterSymbolizer, it uses the Categorize
> function
> and the Interpolate one, not the Recode one.
Section 11.6.4 of the specs
(05-077_Symbology_Encoding_Implementation_Specification.pdf) includes
Recode along with Interpola
On Mon, Jul 4, 2011 at 12:08 PM, Jody Garnett wrote:
> That would be great Andrea; making it actually work fast is required
> before we use it on rasters. I do not know why the code is case insensitive.
>
Working on rasters... hmmm... does not make sense to me, for the raster case
it should be r
That would be great Andrea; making it actually work fast is required before we
use it on rasters. I do not know why the code is case insensitive.
--
Jody Garnett
On Monday, 4 July 2011 at 8:04 PM, Andrea Aime wrote:
> Hi,
> I'm looking into optimizing the recode function for the case where al
Hi,
I'm looking into optimizing the recode function for the case where all keys
and values are literals.
It seems to me the simplest choice is to build a hashmap in the constructor
that maps between keys and values and then use it in the evaluate,
it would be the fastest option available I think.
10 matches
Mail list logo
