“Works for me”, what’s your version collection? pramsey=# select postgis_full_version();
POSTGIS="3.6.0dev 3.5.0-228-g2c8f40113" [EXTENSION] PGSQL="170" GEOS="3.14.0dev-CAPI-1.20.0" PROJ="9.3.0 NETWORK_ENABLED=ON URL_ENDPOINT=https://cdn.proj.org USER_WRITABLE_DIRECTORY=/Users/pramsey/Library/Application Support/proj DATABASE_PATH=/usr/local/share/proj/proj.db" (compiled against PROJ 9.3.0) GDAL="GDAL 3.9.3, released 2024/12/10 (debug build)" LIBXML="2.9.13" LIBJSON="0.17" LIBPROTOBUF="1.4.1" WAGYU="0.5.0 (Internal)" (core procs from "3.6.0dev 3.5.0-186-g2845d3f37" need upgrade) RASTER (raster lib from "3.5.0 3.5.0-211-g202d59f44" need upgrade) (raster procs from "3.5.0 3.5.0" need upgrade) > On Apr 3, 2025, at 2:43 PM, Denis Rykov <ryk...@gmail.com> wrote: > > Is it expected for ST_ClipByBox2D to return NaN coordinates for 'POLYGON Z' > geometries? > > Here is an example. The following SQL query: > > SELECT > ST_AsText( > ST_SimplifyPreserveTopology( > ST_ClipByBox2D( > ST_GeomFromText( > 'POLYGON Z ((9109727.260586184 7235765.611871982 > 0,9109793.397475857 7235780.37911723 0,9109880.551536862 7235799.226171476 > 0,9109912.974452058 7235743.263961895 0,9109949.41754929 7235677.287550647 > 0,9109986.927785816 7235610.625640362 0,9110029.388466287 7235535.445041807 > 0,9110063.152192863 7235481.075809712 0,9110058.528985653 7235478.79780619 > 0,9110066.094453445 7235463.4091950925 0,9110094.816342745 7235418.349932027 > 0,9110128.716586944 7235357.442269537 0,9110170.311475156 7235283.992565744 > 0,9110207.570403146 7235217.92016041 0,9110271.536255063 7235090.866183731 > 0,9110290.55592083 7235042.889807568 0,9110192.389007365 7235014.391634338 > 0,9110169.194172846 7235007.739006326 0,9109951.366830455 7235432.506918687 > 0,9109937.453975307 7235459.0955318 0,9109797.566462943 7235663.435298668 > 0,9109727.260586184 7235765.611871982 0))' > ), > ST_MakeEnvelope( > 9108725.486174684, 7235101.049109253, > 9111416.069569945, 7237791.632504516 > ) > ), > 5 > ) > ); > > fails with the 'Geometry contains invalid coordinates' error, but wrapping > the geometry in ST_Force2D makes it work.
