AnsweredAssumed Answered

What is the proper way to reproject a geometry object using arcpy?

Question asked by eric.eagle on Jun 10, 2020
Latest reply on Jun 10, 2020 by danretired

I am working with a geometry object projected in EPSG 6455 but would like to add its extent to a feature class that is in EPSG 4326.


arcpy.Project_management() wants a feature class as input and output.


Here's what I've got so far (please note that using ZLAS is in no way an endorsement of ZLAS as a proprietary format)


### ZLAS testing ###
zlas = "C:/Data/USGS_LPC_IL_4County_Cook_2017_LAS_15008550_LAS_2019.zlas"

des = arcpy.Describe(zlas)


polys = []

las_extent = [
                [des.extent.YMin, des.extent.XMin],
                [des.extent.YMax, des.extent.XMin],
                [des.extent.YMax, des.extent.XMax],
                [des.extent.YMin, des.extent.XMax]
for coord_ring in las_extent:
    poly = arcpy.Polygon(arcpy.Array([arcpy.Point(*coords) for coords in coord_ring]), arcpy.SpatialReference(6455))

How would I coerce "poly" from 6455 to 4326 without incurring the IO expense of writing feature classes?


The background, if useful: I am plowing through a directory structure with thousands and thousands of point clouds.  The intent is to gather their extents, add them to a global feature class, and then serve that feature class up as a service identifying where all of our holdings are, allowing users to browse our holdings geographically rather than trying to hunt through a directory tree.  Because many of the files are projected differently (some requiring transformation - default transformation being fine as this isn't high precision) I want to harmonize them all into a single GCS feature class.