[PROJ] Migrating to PROJ-9
Even Rouault
even.rouault at spatialys.com
Tue May 30 11:02:20 PDT 2023
It should work at least with the API using proj_factors() since PROJ
8.2. Cf https://github.com/OSGeo/PROJ/pull/2868 (you should likely strip
off the +init= prefix then)
Le 30/05/2023 à 19:56, Nicolas Bellaiche a écrit :
> Ok i hoped your answer would help me understanding why I can't compute
> the scale factor properly but unfortunately it doesn't give me any idea.
> If I instantiate a PJ* with the new syntax (with proj_create), the
> meridional_scale is not computed correctly with proj_factors, while
> with the old "+init=" it still works.
> N.
>
>
> ------------------------------------------------------------------------
> *De: *"Even Rouault" <even.rouault at spatialys.com>
> *À: *"Nicolas Bellaiche" <nicolas.bellaiche at ign.fr>
> *Cc: *"proj" <proj at lists.osgeo.org>
> *Envoyé: *Mardi 30 Mai 2023 19:48:17
> *Objet: *Re: [PROJ] Migrating to PROJ-9
>
>
> Le 30/05/2023 à 19:24, Nicolas Bellaiche a écrit :
>
>
> Ok, i kind of fixed my issue even if that's not all very clear to me.
> One of the things that I can't understand is that outdated "+init="
>
> For instance if I try to use
>
> echo 2d20\'11.5\"E46d47\'59.775\"N |./bin/projIGNF:LAMBE -Swithout
> the +init= syntaxI get
>
>
> proj_create: unrecognized format / unknown name
>
> Rel. 9.1.1, December 1st, 2022
>
> <proj>:
>
> projection initialization failure
>
> cause: Invalid PROJ string syntax
>
> program abnormally terminated
>
>
> Is it normal?
>
> Maybe? the old proj utility works with coordinate operations, actually
> only map projections, but not CRS definitions. In PROJ 4.x era, there
> was no such distinction. IGNF:LAMBE is a CRS definition now.
> Furthermore the proj utility has not been updated to resolve objects
> by code/names. (with the +init= prefix, proj errors out with "can't
> initialize operations that take non-angular input coordinates" for me.
> likely some mismatch between radian vs degree input expectations). I
> presume it could be enhanced to work with projected CRS definitions,
> although it is not totally clear to me what should be the expectations
> regarding input/output axis order.
>
> Even
>
>
> Sincerely,
>
>
> N.B.
>
>
> ------------------------------------------------------------------------
> *De: *"Even Rouault" <even.rouault at spatialys.com>
> <mailto:even.rouault at spatialys.com>
> *À: *"Nicolas Bellaiche" <nicolas.bellaiche at ign.fr>
> <mailto:nicolas.bellaiche at ign.fr>
> *Cc: *"proj" <proj at lists.osgeo.org> <mailto:proj at lists.osgeo.org>
> *Envoyé: *Mardi 30 Mai 2023 14:43:03
> *Objet: *Re: [PROJ] Migrating to PROJ-9
>
> Weird indeed:
>
> and what about trying directly the pipeline with cct :
>
> echo 652760.73700000 7033791.24300000 43.64200000 | cct
> +proj=pipeline +step +inv +proj=lcc +lat_0=46.5 +lon_0=3 +lat_1=44
> +lat_2=49 +x_0=700000 +y_0=6600000 +ellps=GRS80 +step +proj=cart
> +ellps=GRS80
>
> you should get:
>
> 4070327.2514 166078.9321 4891298.6149 inf
>
> Le 30/05/2023 à 14:35, Nicolas Bellaiche a écrit :
>
> Bonjour M. Rouault,
>
> *
> *
>
> Voilà ce que j'obtiens:
>
> *
> *
>
> *PROJ9:*
>
> *
> *
>
> *Release % PROJ_DEBUG=3 projinfo -s IGNF:LAMB93 -t IGNF:RGF93
> -o PROJ*
>
>
>
> pj_open_lib(proj.ini): call
> fopen(/Users/nbellaiche/dev/sd-externs/macosx-clang/proj-9.1.1/nad/proj.ini)
> - succeeded
>
> pj_open_lib(proj.db): call
> fopen(/Users/nbellaiche/dev/sd-externs/macosx-clang/proj-9.1.1/nad/proj.db)
> - succeeded
>
> Candidate operations found: 1
>
> -------------------------------------
>
> Operation No. 1:
>
>
> unknown id, Inverse of LAMBERT-93 + Conversion from RGF93
> geographiques (dms) to RGF93 cartesiennes geocentriques, 0 m,
> World
>
>
> PROJ string:
>
> +proj=pipeline
>
> +step +inv +proj=lcc +lat_0=46.5 +lon_0=3 +lat_1=44 +lat_2=49
> +x_0=700000
>
> +y_0=6600000 +ellps=GRS80
>
> +step +proj=cart +ellps=GRS80
>
>
>
> * Release % echo 652760.73700000 7033791.24300000 43.64200000
> | cs2cs +init=IGNF:LAMB93 +to +init=IGNF:RGF93*
>
> 4070355.05166080.07 4891332.24
>
>
>
> * Release % echo $PROJ_DATA*
>
> /Users/nbellaiche/dev/sd-externs/macosx-clang/proj-9.1.1/nad
>
>
> * Release % echo $PROJ_LIB*
>
> /Users/nbellaiche/dev/sd-externs/macosx-clang/proj-9.1.1/nad
>
>
>
> je ne vois pas de différence. Je n'ai pas le début d'une idée
> de comment identifier le problème...
>
>
> N.
>
>
>
>
> ------------------------------------------------------------------------
> *De: *"Even Rouault" <even.rouault at spatialys.com>
> <mailto:even.rouault at spatialys.com>
> *À: *"proj" <proj at lists.osgeo.org> <mailto:proj at lists.osgeo.org>
> *Envoyé: *Vendredi 26 Mai 2023 21:50:27
> *Objet: *Re: [PROJ] Migrating to PROJ-9
>
> Ah, and you should avoid using the "+init=" prefix. This is a
> deprecated legacy syntax that doesn't enforce axis order of
> the CRS definition (without practical consequence for IGNF:
> CRS since they use "GIS friendly" axis order)
>
> To help debugging this, please provide the output of:
>
> PROJ_DEBUG=3 projinfo -s IGNF:LAMB93 -t IGNF:RGF93 -o PROJ
>
> On my system I get:
>
> pj_open_lib(proj.ini): call
> fopen(/home/even/proj/install-proj-master/share/proj/proj.ini)
> - succeeded
> pj_open_lib(proj.db): call
> fopen(/home/even/proj/install-proj-master/share/proj/proj.db)
> - succeeded
> Candidate operations found: 1
> -------------------------------------
> Operation No. 1:
>
> unknown id, Inverse of LAMBERT-93 + Conversion from RGF93
> geographiques (dms) to RGF93 cartesiennes geocentriques, 0 m,
> World
>
> PROJ string:
> +proj=pipeline
> +step +inv +proj=lcc +lat_0=46.5 +lon_0=3 +lat_1=44
> +lat_2=49 +x_0=700000
> +y_0=6600000 +ellps=GRS80
> +step +proj=cart +ellps=GRS80
>
> Le 26/05/2023 à 21:46, Even Rouault a écrit :
>
> Hi,
>
> I get the expected 4070327.25 166078.93 4891298.61 result
> with proj 6.0.0, 7.2.1, 9.2.0 and master
>
> There's no grid involved in this operation, just inverse
> LCC projection on the RGF93 datum
>
> I also tried to reproduce your issue if the PROJ < 6 IGNF
> file is found in the PROJ resource path(s) (non nominal
> situation - this file is not needed any more, and will
> generally lead to suboptimal results. Its content is
> included in the proj.db file), but even with that I get
> the expected result. So no idea why you get wrong results
>
> Even
>
>
> Le 26/05/2023 à 20:54, Nicolas Bellaiche a écrit :
>
> Hi all,
>
> I'm migrating my C++ code from PROJ4 to PROJ9 and I
> meet a worrying issue with the computation. Some of my
> regress tests don't pass so I've tested with cs2cs and
> gets the same difference:
>
> PROJ4
>
> echo652760.737000007033791.24300000 43.64200000 |
> cs2cs +init=IGNF:LAMB93 +to +init=IGNF:RGF93:
>
> 4070327.25166078.93 4891298.61
>
>
> PROJ7 or PROJ9
>
> echo652760.737000007033791.24300000 43.64200000 |
> cs2cs +init=IGNF:LAMB93 +to +init=IGNF:RGF93:
>
> 4070355.05166080.07 4891332.24
>
>
>
>
> I suspect that PROJ7 and PROJ9 don't read properly the
> geoid file but I'm not sure. Someone can help me
> tracking and fixing this problem?
>
> Thanks,
>
> Nicolas Bellaiche
> Institut Géographique National France
>
> _______________________________________________
> PROJ mailing list
> PROJ at lists.osgeo.org
> https://lists.osgeo.org/mailman/listinfo/proj
>
> --
> http://www.spatialys.com
> My software is free, but my time generally not.
>
>
> _______________________________________________
> PROJ mailing list
> PROJ at lists.osgeo.org
> https://lists.osgeo.org/mailman/listinfo/proj
>
> --
> http://www.spatialys.com
> My software is free, but my time generally not.
>
>
> _______________________________________________
> PROJ mailing list
> PROJ at lists.osgeo.org
> https://lists.osgeo.org/mailman/listinfo/proj
>
> --
> http://www.spatialys.com
> My software is free, but my time generally not.
>
>
> --
> http://www.spatialys.com
> My software is free, but my time generally not.
>
--
http://www.spatialys.com
My software is free, but my time generally not.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.osgeo.org/pipermail/proj/attachments/20230530/7ed64201/attachment-0001.htm>
More information about the PROJ
mailing list