Discussion:
Is kerneldensity_attribute implemented?
Rahkonen Jukka (Tike)
2014-10-23 04:55:20 UTC
Permalink
Hi,

I have tried to use attribute based weights in heatmaps as in RFC 108 example http://mapserver.org/development/rfc/ms-rfc-108.html
processing "KERNELDENSITY_ATTRIBUTE=POP2010"

However, this processing option does not seem to have any affect. It is not used in the MS autotest mapfile and I wonder if it has been implemented at all yet. Does somebody know?

-Jukka Rahkonen-
thomas bonfort
2014-10-23 06:57:09 UTC
Permalink
Jukka,
All the functionality is tested in msautotest, the
KERNELDENSITY_ATTRIBUTE is an error is the RFC. To use an attribute
use STYLE SIZE [attribute] END in your linked vector layer.

--
thomas

On 23 October 2014 06:55, Rahkonen Jukka (Tike)
Post by Rahkonen Jukka (Tike)
Hi,
I have tried to use attribute based weights in heatmaps as in RFC 108 example http://mapserver.org/development/rfc/ms-rfc-108.html
processing "KERNELDENSITY_ATTRIBUTE=POP2010"
However, this processing option does not seem to have any affect. It is not used in the MS autotest mapfile and I wonder if it has been implemented at all yet. Does somebody know?
-Jukka Rahkonen-
_______________________________________________
mapserver-users mailing list
http://lists.osgeo.org/mailman/listinfo/mapserver-users
Rahkonen Jukka (Tike)
2014-10-23 08:21:03 UTC
Permalink
Hi,

Fine, I could make is work with this knowledge. Is it so that the only documentation right now is in the RFC and main documentation page http://mapserver.org/documentation.html does not know about dynamic heatmaps yet? Might be worth having an own entry in the Output section.

I was trying to normalize my source data. There seems to be some limited possibilities for doing that with source layer and kerneldensity layer:
- In source LAYER the attribute to be used for weighting can be selected with "STYLE SIZE [attribute] END. The [attribute] is selected as it stands and no expressions can be used in STYLE - SIZE
- In the kerneldensity LAYER " processing "KERNELDENSITY_NORMALIZATION" can be used for normalizing data by a multiplication factor.
- For more advanced normalization it must be done in the data because for example expression [POPULATION]/[AREA] can't be resolved on-the-fly because expressions are not supported in STYLE - SIZE.

Have I understood right? With some sources of the point layer data there could perhaps be a workaround by computing new field on-the-fly inside LAYER - DATA: "select geometry, population/area as popar from source" and by using then "popar" as size attribute. I made some trials with connnectiontype OGR but it was not trivial to make my selection to work from a shapefile.

-Jukka Rahkonen-
Post by thomas bonfort
Jukka,
All the functionality is tested in msautotest, the KERNELDENSITY_ATTRIBUTE is
an error is the RFC. To use an attribute use STYLE SIZE [attribute] END in your
linked vector layer.
--
thomas
On 23 October 2014 06:55, Rahkonen Jukka (Tike)
Post by Rahkonen Jukka (Tike)
Hi,
I have tried to use attribute based weights in heatmaps as in RFC 108
example http://mapserver.org/development/rfc/ms-rfc-108.html
processing "KERNELDENSITY_ATTRIBUTE=POP2010"
However, this processing option does not seem to have any affect. It is not
used in the MS autotest mapfile and I wonder if it has been implemented at all
yet. Does somebody know?
Post by Rahkonen Jukka (Tike)
-Jukka Rahkonen-
_______________________________________________
mapserver-users mailing list
http://lists.osgeo.org/mailman/listinfo/mapserver-users
Yves Jacolin
2014-10-23 12:04:08 UTC
Permalink
Hello,

If documentation is not uptodate you can either send a PR or create an issue
with some text and information (where? what?).

More information you give, easier and better it is for us.

Y.
Post by Rahkonen Jukka (Tike)
Hi,
Fine, I could make is work with this knowledge. Is it so that the only
documentation right now is in the RFC and main documentation page
http://mapserver.org/documentation.html does not know about dynamic
heatmaps yet? Might be worth having an own entry in the Output section.
I was trying to normalize my source data. There seems to be some limited
possibilities for doing that with source layer and kerneldensity layer: -
In source LAYER the attribute to be used for weighting can be selected with
"STYLE SIZE [attribute] END. The [attribute] is selected as it stands and
no expressions can be used in STYLE - SIZE - In the kerneldensity LAYER "
processing "KERNELDENSITY_NORMALIZATION" can be used for normalizing data
by a multiplication factor. - For more advanced normalization it must be
done in the data because for example expression [POPULATION]/[AREA] can't
be resolved on-the-fly because expressions are not supported in STYLE -
SIZE.
Have I understood right? With some sources of the point layer data there
could perhaps be a workaround by computing new field on-the-fly inside
LAYER - DATA: "select geometry, population/area as popar from source" and
by using then "popar" as size attribute. I made some trials with
connnectiontype OGR but it was not trivial to make my selection to work
from a shapefile.
-Jukka Rahkonen-
Post by thomas bonfort
Jukka,
All the functionality is tested in msautotest, the KERNELDENSITY_ATTRIBUTE
is an error is the RFC. To use an attribute use STYLE SIZE [attribute]
END in your linked vector layer.
--
thomas
On 23 October 2014 06:55, Rahkonen Jukka (Tike)
Post by Rahkonen Jukka (Tike)
Hi,
I have tried to use attribute based weights in heatmaps as in RFC 108
example http://mapserver.org/development/rfc/ms-rfc-108.html
processing "KERNELDENSITY_ATTRIBUTE=POP2010"
However, this processing option does not seem to have any affect. It is not
used in the MS autotest mapfile and I wonder if it has been implemented at
all yet. Does somebody know?
Post by Rahkonen Jukka (Tike)
-Jukka Rahkonen-
_______________________________________________
mapserver-users mailing list
http://lists.osgeo.org/mailman/listinfo/mapserver-users
_______________________________________________
mapserver-users mailing list
http://lists.osgeo.org/mailman/listinfo/mapserver-users
--
Responsable Formation et Support
Camptocamp France SAS
Savoie Technolac, BP 352
73377 Le Bourget du Lac, Cedex
Tel (France) : +33 4 79 26 57 98
Tel (Suisse) : 021 619 10 43
Mob. : +33 6 18 75 42 21
Fax : 04 79 70 15 81
Mail : ***@camptocamp.com
http://www.camptocamp.com
Rahkonen Jukka (Tike)
2014-10-23 12:58:04 UTC
Permalink
Hi,

I made an issue to Github. By browsing through the list of RFCs there may be also some other features in Mapserver which are best or only documented in corresponding RFCs. That is OK as an initial documentation but if things are a 'changing and usage is perhaps no more exactly as it was written to RFC. It looks like there is usually no considerations in RFCs about how and where to document the fine new features for the users.

-Jukka Rahkonen-
Post by Yves Jacolin
Hello,
If documentation is not uptodate you can either send a PR or create an issue
with some text and information (where? what?).
More information you give, easier and better it is for us.
Y.
Post by Rahkonen Jukka (Tike)
Hi,
Fine, I could make is work with this knowledge. Is it so that the only
documentation right now is in the RFC and main documentation page
http://mapserver.org/documentation.html does not know about dynamic
heatmaps yet? Might be worth having an own entry in the Output section.
I was trying to normalize my source data. There seems to be some
limited possibilities for doing that with source layer and
kerneldensity layer: - In source LAYER the attribute to be used for
weighting can be selected with "STYLE SIZE [attribute] END. The
[attribute] is selected as it stands and no expressions can be used in STYLE -
SIZE - In the kerneldensity LAYER "
Post by Rahkonen Jukka (Tike)
processing "KERNELDENSITY_NORMALIZATION" can be used for normalizing
data by a multiplication factor. - For more advanced normalization it
must be done in the data because for example expression
[POPULATION]/[AREA] can't be resolved on-the-fly because expressions
are not supported in STYLE - SIZE.
Have I understood right? With some sources of the point layer data
there could perhaps be a workaround by computing new field on-the-fly
inside LAYER - DATA: "select geometry, population/area as popar from
source" and by using then "popar" as size attribute. I made some
trials with connnectiontype OGR but it was not trivial to make my
selection to work from a shapefile.
-Jukka Rahkonen-
Post by thomas bonfort
Jukka,
All the functionality is tested in msautotest, the
KERNELDENSITY_ATTRIBUTE is an error is the RFC. To use an attribute
use STYLE SIZE [attribute] END in your linked vector layer.
--
thomas
On 23 October 2014 06:55, Rahkonen Jukka (Tike)
Post by Rahkonen Jukka (Tike)
Hi,
I have tried to use attribute based weights in heatmaps as in RFC
108 example http://mapserver.org/development/rfc/ms-rfc-108.html
processing "KERNELDENSITY_ATTRIBUTE=POP2010"
However, this processing option does not seem to have any affect.
It is not
used in the MS autotest mapfile and I wonder if it has been
implemented at all yet. Does somebody know?
Post by Rahkonen Jukka (Tike)
-Jukka Rahkonen-
_______________________________________________
mapserver-users mailing list
http://lists.osgeo.org/mailman/listinfo/mapserver-users
_______________________________________________
mapserver-users mailing list
http://lists.osgeo.org/mailman/listinfo/mapserver-users
--
Responsable Formation et Support
Camptocamp France SAS
Savoie Technolac, BP 352
73377 Le Bourget du Lac, Cedex
Tel (France) : +33 4 79 26 57 98
Tel (Suisse) : 021 619 10 43
Mob. : +33 6 18 75 42 21
Fax : 04 79 70 15 81
http://www.camptocamp.com
Loading...