sis-commits mailing list archives: April 2019

Site index · List index
Message listThread · Author · Date
Martin Desruisseaux (JIRA) [jira] [Created] (SIS-450) Sinusoidal projection Sat, 20 Apr, 09:59
Martin Desruisseaux (JIRA) [jira] [Resolved] (SIS-450) Sinusoidal projection Thu, 25 Apr, 13:40
Martin Desruisseaux (JIRA) [jira] [Assigned] (SIS-227) American Polyconic (EPSG:9818) Tue, 30 Apr, 12:06
Martin Desruisseaux (JIRA) [jira] [Resolved] (SIS-227) American Polyconic (EPSG:9818) Tue, 30 Apr, 12:07
Martin Desruisseaux (JIRA) [jira] [Updated] (SIS-227) American Polyconic (EPSG:9818) Tue, 30 Apr, 12:13
Martin Desruisseaux (JIRA) [jira] [Updated] (SIS-227) American Polyconic (EPSG:9818) Tue, 30 Apr, 12:14
Martin Desruisseaux (JIRA) [jira] [Created] (SIS-451) Verify map projection derivative by comparaison with Snyder terms Tue, 30 Apr, 12:57
Martin Desruisseaux (JIRA) [jira] [Updated] (SIS-370) Improve CoordinateOperation WKT for axis changes and Molodensky parameters Tue, 30 Apr, 14:11
desruisse...@apache.org [sis] branch geoapi-4.0 updated: Refactor WraparoundAdjustment in a way that makes easier to share code between the method working on Envelope and the method working on DirectPosition. Sun, 07 Apr, 19:07
desruisse...@apache.org [sis] branch geoapi-4.0 updated: Make GridDerivation.subgrid(…) consistent with GridDerivation.slice(…) regarding IncompleteGridGeometryException being thrown if `areaOfInterest` has a CRS but `GridGeometry` has none. Mon, 08 Apr, 08:12
desruisse...@apache.org [sis] branch geoapi-4.0 updated: If correction for ±180° meridian crossing results in a GridGeometry well outside its original longitude range, apply a 360° translation on the whole grid. Mon, 08 Apr, 12:22
desruisse...@apache.org [sis] branch geoapi-4.0 updated: Add documentation and checks against integer overflow (ArithmeticException). Make ConvertedGridCoverage package-private and declare ConvertedColorModel as compatible with ConvertedSampleModel only. Avoid the use of MathTransforms.compound(…) since in this particular case an ordinary loop will be more efficient. Mon, 08 Apr, 19:00
desruisse...@apache.org [sis] branch geoapi-4.0 updated: Internal class renaming. Tue, 09 Apr, 07:49
desruisse...@apache.org [sis] branch geoapi-4.0 updated: Fetching file extension shall work in the case where the given `storage` argument is already an instance of StorageConnector. Tue, 09 Apr, 09:24
desruisse...@apache.org [sis] branch geoapi-4.0 updated: Add comment explaining better how the scale factors are computed in `GridDerivation.subgrid(…)`. Tue, 09 Apr, 12:46
desruisse...@apache.org [sis] branch geoapi-4.0 updated: Resolve name collision in RasterResource.identifier when two netCDF variables use the same "standard_name" attribute value. Tue, 09 Apr, 14:43
desruisse...@apache.org [sis] branch geoapi-4.0 updated: Fix an IllegalArgumentException when creating one-banded image of type DataBuffer.TYPE_INT. Remove ScaledColorSpace from public API (provide a static method instead). Remove ColorModelFactory.getNumBands(ColorModel) method. Wed, 10 Apr, 09:02
desruisse...@apache.org [sis] branch geoapi-4.0 updated (3b01827 -> 6092c0a) Fri, 12 Apr, 18:22
desruisse...@apache.org [sis] 01/03: Revert commit a685c8a1bf0b412a544713f1ceda9749a995a5d6 (compute linear regression by minimizing errors in grid indices instead than geospatial coordinates). The reason is because the "compute linear regression in reverse way" approach brings more complexity and is not sufficient anyway; we still needs a more sophisticated iterative algorithm in InterpolatedTransform.Inverse. Fri, 12 Apr, 18:22
desruisse...@apache.org [sis] 02/03: Add explanation about the "compute linear regression in reverse way" approach that we reverted in previous commit, so if we want to try it again we know which commit to revert. Fri, 12 Apr, 18:22
desruisse...@apache.org [sis] 03/03: First draft of an InterpolatedTransform algorithm using derivative (Jacobian matrix) for deciding in which direction to compensate for errors. Not yet enabled. Fri, 12 Apr, 18:22
desruisse...@apache.org [sis] branch geoapi-4.0 updated: Refactor the way we use ParameterDescriptorGroup, ParameterValueGroup and "grid to target" transform in GridDatumShift in such a way that we can keep GridDatumShiftFile internal to NADCON and NTv2 transformations. This avoid the confusion declaration of ResidualGrid as a subtype of DatumShiftGridFile. Also moved the DatumShiftGrid.normalizedToGridX/Y methods as package-private methods in DatumShiftTransform; those methods should never have been public, since their relationship with `coordi [...] Sat, 13 Apr, 16:07
desruisse...@apache.org [sis] branch geoapi-4.0 updated: Netcdf file opened with the UCAR library should use NetcdfDataset.Enhance.CoordSystems at construction time. Enhancements just before requesting coordinate systems is sometime too late. The symptom we observed was that if metadata were requested before the resources, no grid resources were found because the CS were not found. Invoking NetcdfDataset.enhance(…) at construction time fix that. Sat, 13 Apr, 22:04
desruisse...@apache.org [sis] branch geoapi-4.0 updated: Enable InterpolatedTransform.Inverse method based on Jacobian matrix (but still needs more test). Sun, 14 Apr, 17:36
desruisse...@apache.org [sis] branch geoapi-4.0 updated: Calculation of derivative outside the grid must be consistent with the way values are extrapolated by `interpolateInCell`. Mon, 15 Apr, 10:55
desruisse...@apache.org [sis] branch geoapi-4.0 updated: Add a "Satellite ground track" transform (internal only) for trying to make GCOM data a little bit more linear before to let LocalizationGridBuilder to its job. Mon, 15 Apr, 20:10
desruisse...@apache.org [sis] branch geoapi-4.0 updated (6fd90af -> 8b63089) Tue, 16 Apr, 12:51
desruisse...@apache.org [sis] 01/03: Tuning of number of itetations and accuracy threshold. Tue, 16 Apr, 12:51
desruisse...@apache.org [sis] 02/03: Minor cleanup of debugging code. Tue, 16 Apr, 12:51
desruisse...@apache.org [sis] 03/03: Store localization grids using single precision instead of double precision for reducing memory consumption. We presume that single precision is sufficient because the values stored in the `offsets` array are differences, not absolute coordinates. The absolute coordinates will be computed by adding `double` values to those offsets. Tue, 16 Apr, 12:51
desruisse...@apache.org [sis] branch geoapi-4.0 updated: Reduce the amount of copy-and-paste in LinearIteratorTest by leveraging the tests provided by DefaultIteratorTest. Tue, 16 Apr, 19:06
desruisse...@apache.org [sis] branch geoapi-4.0 updated: Use StrictMath in test classes as a matter of principle. Tue, 16 Apr, 19:06
desruisse...@apache.org [sis] branch geoapi-4.0 updated: Do not return NaN for "no convergence" even in the case of extrapolations, because tests have shown that the `grid.isCellInGrid(x, y)` may return `false` (i.e. the point seems to be outside the grid) even when in reality it should have been inside. In other words, we can not be sure that an "extrapolation" is really outside the grid. The algorithm may have fail to converge and think that the point is outside even when the point was deeply inside (not just on the border) if curvatures in loc [...] Wed, 17 Apr, 08:40
desruisse...@apache.org [sis] branch geoapi-3.1 updated (d416de4 -> 93dca74) Wed, 17 Apr, 08:49
desruisse...@apache.org [sis] 01/01: Merge branch 'geoapi-4.0' into geoapi-3.1. This include mostly netCDF work, with sis-raster improvements in support. Wed, 17 Apr, 08:49
desruisse...@apache.org [sis] branch master updated (002121a -> e55b128) Wed, 17 Apr, 09:34
desruisse...@apache.org [sis] 01/01: Merge branch 'geoapi-3.1'. The main topic is more development in netCDF reader, together with improvements in other modules in support of netCDF needs. Wed, 17 Apr, 09:34
desruisse...@apache.org [sis] branch geoapi-4.0 updated: Edit `GridCoverageResource` contract by allowing `getGridGeometry()` and `getSampleDimensions()` to be slightly different than the corresponding `GridCoverage` properties. Thu, 18 Apr, 09:37
desruisse...@apache.org [sis] branch geoapi-4.0 updated: If a netCDF variable contains "flag_names" or "flag_meanings" attribute, create categories for enumeration values instead than for the range of quantitative values. Thu, 18 Apr, 16:59
desruisse...@apache.org [sis] branch geoapi-4.0 updated: First draft of "Sinusoidal equal-area" projection, also known as "Sanson-Flamsteed". This commit put only the spherical formulas. Ellipsoidal formulas will come later. https://issues.apache.org/jira/browse/SIS-450 Sat, 20 Apr, 21:30
desruisse...@apache.org [sis] branch geoapi-4.0 updated (33b6d27 -> 14efbad) Sat, 20 Apr, 21:34
desruisse...@apache.org [sis] 01/01: First draft of "Sinusoidal equal-area" projection, also known as "Sanson-Flamsteed". This commit put only the spherical formulas. Ellipsoidal formulas will come later. https://issues.apache.org/jira/browse/SIS-450 Sat, 20 Apr, 21:34
desruisse...@apache.org [sis] branch geoapi-4.0 updated: Add skeleton base class for meridional distance calculation on ellipsoid. Add formulas to be used as reference in the test class (the formula actually used by implementation may differ). https://issues.apache.org/jira/browse/SIS-450 Mon, 22 Apr, 15:59
desruisse...@apache.org svn commit: r1858015 - /sis/analysis/ Tue, 23 Apr, 17:33
desruisse...@apache.org svn commit: r1858016 - /sis/analysis/Map projection formulas.ods Tue, 23 Apr, 17:41
desruisse...@apache.org [sis] branch geoapi-4.0 updated: Implement MeridionalDistance.meridianArc(…). Tue, 23 Apr, 17:45
desruisse...@apache.org [sis] branch geoapi-4.0 updated: Adjust the computation of "gridToCRS" in pixel center when derivating from another grid geometry. Wed, 24 Apr, 11:26
desruisse...@apache.org svn commit: r1858041 - /sis/analysis/README.html Wed, 24 Apr, 12:31
desruisse...@apache.org [sis] branch geoapi-4.0 updated (8158557 -> d2c0625) Wed, 24 Apr, 13:57
desruisse...@apache.org [sis] 01/03: Improve comments about the adjustment done for PixelInCell.CELL_CENTER when deriving a GridGeometry from another one. Wed, 24 Apr, 13:57
desruisse...@apache.org [sis] 02/03: Rename an "ellipse" boolean parameter as "ellipsoidal". Wed, 24 Apr, 13:57
desruisse...@apache.org [sis] 03/03: Fix derivative calculation and add tests. Wed, 24 Apr, 13:57
desruisse...@apache.org [sis] branch geoapi-4.0 updated: Implement MeridionalDistance.inverse(double). Wed, 24 Apr, 18:53
desruisse...@apache.org svn commit: r1858068 - /sis/analysis/Map projection formulas.ods Wed, 24 Apr, 19:00
desruisse...@apache.org [sis] branch geoapi-4.0 updated: Complete the tests for MeridionalDistance inverse method. Thu, 25 Apr, 08:06
desruisse...@apache.org [sis] branch geoapi-4.0 updated: Rename "MeridionalDistanceBased" class as "MeridianArcBased" because online dictionaries seem to define "meridional distance" as "the distance or departure from the meridian; the easting or westing", which is not what we mean. EPSG guidance note and Snyder do not use "meridional distance" expression neither for the quantity computed here. Thu, 25 Apr, 09:42
desruisse...@apache.org [sis] branch geoapi-4.0 updated: Implement ellipsoidal formulas for the sinusoidale projection. This commit completes https://issues.apache.org/jira/browse/SIS-450 Thu, 25 Apr, 13:39
desruisse...@apache.org [sis] branch geoapi-4.0 updated: Partial work toward parsing netCDF "grid_mapping" attributes. Thu, 25 Apr, 22:41
desruisse...@apache.org [sis] branch geoapi-4.0 updated: Add inference of "grid to CRS" transform from netCDF attributes associated to "grid_mapping" variable. Fri, 26 Apr, 10:36
desruisse...@apache.org [sis] branch geoapi-4.0 updated: More complete parsing of "grid_mapping" parameters specified by CF-conventions. Fri, 26 Apr, 16:08
desruisse...@apache.org [sis] branch geoapi-4.0 updated (b638306 -> 3dbcae8) Sat, 27 Apr, 16:14
desruisse...@apache.org [sis] 01/02: Consolidation of parsing of "grid_mapping" attributes in netCDF file: - Change order of some methods in Convention class for grouping "domain" or "range" aspects together. - Take in account the "latitude_longitude" pseudo-projection special case. - Add a Convention.defaultHorizontalCRS(boolean) method in replacement of CRSBuilder.DEFAULT constant. - Replace some in-line strings for attribute name by constants. Sat, 27 Apr, 16:14
desruisse...@apache.org [sis] 02/02: Rename org.apache.sis.metadata.fra module as org.apache.sis.profile.france. The reason is that national profiles may be on other aspects than metadata. Sat, 27 Apr, 16:14
desruisse...@apache.org [sis] branch geoapi-4.0 updated: A CSAuthorityFactory containing at least the coordinate system EPSG::4400 should be available even if there is no connection to the EPSG database. This is required for some data stores like netCDF. The CSAuthorityFactory interface is added on EPSGFactoryFallback. Sun, 28 Apr, 15:23
desruisse...@apache.org [sis] branch geoapi-4.0 updated: Apply to ConformalProjection the same Snyder 3-35 formulas (trigonometric identities) than we did for MeridianArcBased. Include the coefficients in the serialization, since recomputing values after deserialization may not produce the exact same results than the values computed at construction time. Since we apply a Snyder formula now, those fields shouls be more stable than before. Sun, 28 Apr, 22:22
desruisse...@apache.org [sis] branch geoapi-4.0 updated (851bd10 -> 5b271a0) Mon, 29 Apr, 16:34
desruisse...@apache.org [sis] 01/03: Cleanup in test and benchmark methods. Mon, 29 Apr, 16:34
desruisse...@apache.org [sis] 02/03: Apply to EqualAreaProjection the same Snyder 3-35 formulas (trigonometric identities) than we did for MeridianArcBased. Include coefficients and other fields in serialization for the same reason than ConformalProjection (simplicity, stability). During this work, we identified a bug in the series expansion of φ(β) where the last term was multiplied by e⁸ instead of e⁶. Fixing that bug bring the series expansion to a precision where we no longer need iterations at least with WGS84 ellipsoid. Mon, 29 Apr, 16:34
desruisse...@apache.org [sis] 03/03: Include MeridianArcBased fields in serialization for the same simplicity and numerical stability reasons than ConformalProjection and EqualAreaProjection. Inconvenient is that the serialization is more at risk to break if the formulas implemented in those case are improved. However MathTransform serialization should not be used for long-term storage. It is okay for RMI purpose between JVMs running the same version of SIS. Mon, 29 Apr, 16:34
desruisse...@apache.org svn commit: r1858383 - /sis/analysis/Map projection formulas.ods Mon, 29 Apr, 16:51
desruisse...@apache.org svn commit: r1858419 - /sis/analysis/Map projection formulas.ods Tue, 30 Apr, 10:14
desruisse...@apache.org [sis] branch geoapi-4.0 updated: Port the "American Polyconic" projection (EPSG:9818) from Geotk as a side effect of the work on the Sinudoisal projection (those two projections share a common basis coded in MericianArcBased class) and cleanup some tests based on the experience gained by the addition of those two projections. Tue, 30 Apr, 11:20
desruisse...@apache.org [sis] branch geoapi-4.0 updated (8ca9f9a -> 9dabf8e) Tue, 30 Apr, 14:11
desruisse...@apache.org [sis] 01/02: Remove unnecessary "throws ProjectionException" in Spherical private classes. Tue, 30 Apr, 14:11
desruisse...@apache.org [sis] 02/02: Replace "spatio-temporal" by "spatiotemporal" for more uniform spelling. Tue, 30 Apr, 14:11
desruisse...@apache.org [sis] branch geoapi-4.0 updated: Remove SimpleIdentifier.toWKT() because not needed anymore, and document why PropertyInformation extends SimpleIdentifier. Tue, 30 Apr, 14:16
desruisse...@apache.org [sis] branch geoapi-3.1 updated (93dca74 -> 2751173) Tue, 30 Apr, 14:25
desruisse...@apache.org [sis] 01/01: Merge branch 'geoapi-4.0' into geoapi-3.1. Contains work on netCDF and addition of two map projections. Tue, 30 Apr, 14:25
desruisse...@apache.org [sis] branch master updated (e55b128 -> 4610eab) Tue, 30 Apr, 14:35
desruisse...@apache.org [sis] 01/01: Merge branch 'geoapi-3.1'. Contains work on netCDF reader and addition of two map projections. Tue, 30 Apr, 14:35
desruisse...@apache.org [sis] branch geoapi-4.0 updated (b280070 -> 24fcb0f) Tue, 30 Apr, 16:32
desruisse...@apache.org [sis] 01/02: Fix a thread-safety hole. Tue, 30 Apr, 16:32
desruisse...@apache.org [sis] 02/02: Implements GridEnvelope and GridCoordinates following resolution of https://github.com/opengeospatial/geoapi/issues/36 Tue, 30 Apr, 16:32
jso...@apache.org [sis] branch geoapi-4.0 updated: Coverage : add forConvertedValues method Mon, 08 Apr, 09:56
Message listThread · Author · Date
Box list
Nov 201996
Oct 2019189
Sep 2019234
Aug 2019124
Jul 201926
Jun 2019194
May 2019142
Apr 201984
Mar 201975
Feb 201999
Jan 201997
Dec 201890
Nov 201894
Oct 201891
Sep 201816
Aug 201897
Jul 2018130
Jun 2018306
May 201888
Apr 201889
Mar 201868
Feb 2018209
Jan 2018110
Dec 201754
Nov 2017242
Oct 2017174
Sep 201785
Aug 201788
Jul 2017120
Jun 201773
May 2017180
Apr 2017203
Mar 2017105
Feb 201782
Jan 2017192
Dec 2016115
Nov 2016120
Oct 2016113
Sep 2016124
Aug 2016133
Jul 2016167
Jun 201659
May 2016167
Apr 2016126
Mar 2016175
Feb 2016186
Jan 2016160
Dec 2015125
Nov 2015162
Oct 2015288
Sep 2015241
Aug 201567
Jul 2015320
Jun 2015156
May 2015144
Apr 2015133
Mar 2015148
Feb 2015228
Jan 2015164
Dec 201463
Nov 201481
Oct 2014146
Sep 2014114
Aug 201471
Jul 201484
Jun 2014148
May 2014128
Apr 2014200
Mar 201487
Feb 2014142
Jan 2014155
Dec 2013179
Nov 201391
Oct 2013137
Sep 2013139
Aug 2013290
Jul 2013376
Jun 2013291
May 2013219
Apr 2013216
Mar 2013177
Feb 2013145
Jan 201342
Dec 2012164
Nov 201296
Oct 2012200
Sep 2012139
Aug 201230
Jul 201297
May 201210
Apr 201250
Mar 201255
Feb 201215
Aug 201110
Feb 20112
Jan 20113
Nov 2010101
Sep 20102
Aug 201020
Jul 201023
May 20101
Apr 201016
Mar 201012