Start a new topic

Geo-Specific Model Naming Convensions

Original Post by: Icewynd Mon Jun 14 19:21:03 2010

In section in the CDB Spec it specifies \ as the naming convension for the Geo-Specific models within a zip file.

More specifically it mentions NAME as FACC_FSC_MODL where MODL is 32-character user defined model string name. The question I have is does this NAME have to be the same across all levels of detail or can it vary? So if I had a model that has 5 levels of detail lets say L6 - L10 does this NAME need to be the same across all those levels of detail?

If it does not have to be the same across all levels of detail how do you logically connect the model between levels of detail? I am hoping that this NAME needs to be the same.

Any help would be appreciated.


Original Post by: RyanFranz Mon Jun 14 22:34:18 2010

My understanding of the CDB spec is that the model NAME portion of the filename does not change across levels of detail. And the databases that we have used do the same thing.

But I don't work for Presagis, so it is just my reading and usage of the spec. :-)

Original Post by: David.Nadeau Wed Jun 16 00:08:57 2010

Ryan has a good understanding of the specification, which would be expected based on your CDB integration :)

The intent of the specification is to preserve the name across LODs to allow a logical connection between models.

Original Post by: kraj Thu Apr 3 00:08:08 2014

There doesn't appear to be a good way to figure out the file specification of a model from the class attributes. The specification states,

The linkage is made through point-feature attributes which together provide the information needed by client-devices to locate the Model from the appropriate Dataset at the appropriate level-of-detail. The following feature attributes provide the necessary linkage:

FACC-FSC: Feature code and Subcode

MODL : Model Name

MODT: Model Type

MLOD: Model Level-of-Detail

but those fields do not appear to be enough information since, for instance, the U and R values are not specified. How can one determine the appropriate sub-directory a model is located at? For instance, here is a sample record from a file named N38W105_D100_S001_T002_L04_U13_R4.dbf:


FSC: 0

MODL: fbe6eb5f


There are no (obvious) attributes that can be used to link the data from LOD=4;U=13;R=4 to a model with LOD=2 (which can't possibly have a U value of 13). This model happens to live in 300_GSModelGeometry/L02/U3/

What am I missing? Thanks.

Original Post by: B. Leclerc Thu Apr 3 14:06:58 2014

You compute the U/R coordinate of the model from that of the feature, using this simple algo:

U/R(model) = U/R(feature) >> (LOD(feature) - MLOD)

Stated differently, you obtain the U/R coordinate of the model by shifting (dividing by a power of 2) the U/R coordinate of the feature by the difference in LODs between the feature and the model.

That is why the value of MLOD can't exceed the LOD of the Vector (the Shapefile) referring to the model... as indicated in the description of MLOD in section

Original Post by: kraj Thu Apr 3 16:37:38 2014

Thanks much B. Leclerc.

Original Post by: klarakos Fri Aug 1 13:58:04 2014

This is very usefully for me

Login to post a comment