Author: desruisseaux
Date: Sun Aug 2 21:59:35 2015
New Revision: 1693832
URL: http://svn.apache.org/r1693832
Log:
Minor documentation (no code change).
Modified:
sis/branches/JDK8/core/sis-referencing/src/main/java/org/apache/sis/referencing/operation/projection/Initializer.java
sis/branches/JDK8/core/sis-referencing/src/main/java/org/apache/sis/referencing/operation/projection/LambertConicConformal.java
sis/branches/JDK8/core/sis-referencing/src/main/java/org/apache/sis/referencing/operation/projection/Mercator.java
sis/branches/JDK8/core/sis-referencing/src/main/java/org/apache/sis/referencing/operation/projection/NormalizedProjection.java
Modified:
sis/branches/JDK8/core/sis-referencing/src/main/java/org/apache/sis/referencing/operation/projection/Initializer.java
URL:
http://svn.apache.org/viewvc/sis/branches/JDK8/core/sis-referencing/src/main/java/org/apache/sis/referencing/operation/projection/Initializer.java?rev=1693832&r1=1693831&r2=1693832&view=diff
==============================================================================
---
sis/branches/JDK8/core/sis-referencing/src/main/java/org/apache/sis/referencing/operation/projection/Initializer.java
[UTF-8] (original)
+++
sis/branches/JDK8/core/sis-referencing/src/main/java/org/apache/sis/referencing/operation/projection/Initializer.java
[UTF-8] Sun Aug 2 21:59:35 2015
@@ -37,13 +37,22 @@ import static org.apache.sis.internal.ut
* Helper class for map projection constructions, providing formulas normally
needed only at construction time.
* Since map projection constructions should not happen very often, we afford
using some double-double arithmetic.
* The main intend is not to provide more accurate coordinate conversions
(while it may be a nice side-effect),
- * but to improve the result of concatenations of (de)normalization matrices
with the matrices of other transforms,
- * as found in transformation chains.
+ * but to improve the result of matrix multiplications when the map projection
is part of a more complex chain
+ * of transformations. More specifically we want to be able:
+ *
+ * <ul>
+ * <li>To convert degrees to radians, than back to degrees and find the
original value.</li>
+ * <li>To convert axis length (optionally with flattening factor) to
excentricity, then back
+ * to axis length and find the original value.</li>
+ * </ul>
+ *
+ * This has visible effects on WKT formatting among others, but also in our
capability to detect simplification
+ * opportunities in relatively complex chains of transformations.
*
* <p>As a general rule, we stop storing result with double-double precision
after the point where we need
* transcendental functions (sine, logarithm, <i>etc.</i>), since we do not
have double-double versions of
* those functions. Digits after the {@code double} part are usually not
significant in such cases, except
- * in some relatively rare scenarios like 1 ± (a result much smaller than
1).</p>
+ * in some relatively rare scenarios like 1 ± x where <var>x</var> is much
smaller than 1.</p>
*
* @author Martin Desruisseaux (Geomatys)
* @since 0.6
@@ -142,6 +151,12 @@ final class Initializer {
*/
isIvfDefinitive = false;
}
+ /*
+ * The ellipsoid parameters (a, b or ivf) are assumed accurate in
base 10 rather than in base 2,
+ * because they are defined by authorities. For example the
semi-major axis length of the WGS84
+ * ellipsoid is equal to exactly 6378137 metres by definition of
that ellipsoid. The DoubleDouble
+ * constructor applies corrections for making those values more
accurate in base 10 rather than 2.
+ */
if (isIvfDefinitive) {
final DoubleDouble f = new
DoubleDouble(parameters.parameter(Constants.INVERSE_FLATTENING).doubleValue());
f.inverseDivide(1,0);
@@ -178,11 +193,20 @@ final class Initializer {
k.inverseDivide(b, 0);
}
}
- context.normalizeGeographicInputs(λ0);
+ /*
+ * Scale factor is assumed more accurate in base 10 than in base 2 for
the same reason than for the
+ * ellipsoid parameters (i.e. is a value given by authority as part of
map projection definition).
+ * Again, DoubleDouble constructor will take care of computing a
correction.
+ */
final ParameterDescriptor<? extends Number> scaleFactor =
roles.get(ParameterRole.SCALE_FACTOR);
if (scaleFactor != null) {
k.multiply(getAndStore(scaleFactor));
}
+ /*
+ * Set meridian rotation, scale factor, false easting and false
northing parameter values
+ * in the (de)normalization matrices.
+ */
+ context.normalizeGeographicInputs(λ0);
final MatrixSIS denormalize = context.getMatrix(false);
denormalize.convertAfter(0, k, new DoubleDouble(fe));
denormalize.convertAfter(1, k, new DoubleDouble(fn));
Modified:
sis/branches/JDK8/core/sis-referencing/src/main/java/org/apache/sis/referencing/operation/projection/LambertConicConformal.java
URL:
http://svn.apache.org/viewvc/sis/branches/JDK8/core/sis-referencing/src/main/java/org/apache/sis/referencing/operation/projection/LambertConicConformal.java?rev=1693832&r1=1693831&r2=1693832&view=diff
==============================================================================
---
sis/branches/JDK8/core/sis-referencing/src/main/java/org/apache/sis/referencing/operation/projection/LambertConicConformal.java
[UTF-8] (original)
+++
sis/branches/JDK8/core/sis-referencing/src/main/java/org/apache/sis/referencing/operation/projection/LambertConicConformal.java
[UTF-8] Sun Aug 2 21:59:35 2015
@@ -269,7 +269,7 @@ public class LambertConicConformal exten
final double t1 = expOfNorthing(φ1, excentricity*sinφ1);
/*
* Compute n = (ln m₁ – ln m₂) / (ln t₁ – ln t₂), which we rewrite as
ln(m₁/m₂) / ln(t₁/t₂)
- * since division is less at risk of precision lost than subtraction.
Note that this equation
+ * for reducing the amount of calls to the logarithmic function. Note
that this equation
* tends toward 0/0 if φ₁ ≈ φ₂, which force us to do a special check
for the SP1 case.
*/
if (abs(φ1 - φ2) >= ANGULAR_TOLERANCE) { // Should be 'true' for 2SP
case.
Modified:
sis/branches/JDK8/core/sis-referencing/src/main/java/org/apache/sis/referencing/operation/projection/Mercator.java
URL:
http://svn.apache.org/viewvc/sis/branches/JDK8/core/sis-referencing/src/main/java/org/apache/sis/referencing/operation/projection/Mercator.java?rev=1693832&r1=1693831&r2=1693832&view=diff
==============================================================================
---
sis/branches/JDK8/core/sis-referencing/src/main/java/org/apache/sis/referencing/operation/projection/Mercator.java
[UTF-8] (original)
+++
sis/branches/JDK8/core/sis-referencing/src/main/java/org/apache/sis/referencing/operation/projection/Mercator.java
[UTF-8] Sun Aug 2 21:59:35 2015
@@ -246,6 +246,11 @@ public class Mercator extends ConformalP
denormalize.convertBefore(0, k0, null);
denormalize.convertBefore(1, k0, null);
if (λ0 != 0) {
+ /*
+ * Use double-double arithmetic here for consistency with the work
done in the normalization matrix.
+ * The intend is to have exact value at 'double' precision when
computing Matrix.invert(). Note that
+ * there is no such goal for other parameters computed from sine
or consine functions.
+ */
final DoubleDouble offset = DoubleDouble.createDegreesToRadians();
offset.multiply(-λ0);
denormalize.convertBefore(0, null, offset);
Modified:
sis/branches/JDK8/core/sis-referencing/src/main/java/org/apache/sis/referencing/operation/projection/NormalizedProjection.java
URL:
http://svn.apache.org/viewvc/sis/branches/JDK8/core/sis-referencing/src/main/java/org/apache/sis/referencing/operation/projection/NormalizedProjection.java?rev=1693832&r1=1693831&r2=1693832&view=diff
==============================================================================
---
sis/branches/JDK8/core/sis-referencing/src/main/java/org/apache/sis/referencing/operation/projection/NormalizedProjection.java
[UTF-8] (original)
+++
sis/branches/JDK8/core/sis-referencing/src/main/java/org/apache/sis/referencing/operation/projection/NormalizedProjection.java
[UTF-8] Sun Aug 2 21:59:35 2015
@@ -410,7 +410,7 @@ public abstract class NormalizedProjecti
NormalizedProjection(final Initializer initializer) {
context = initializer.context;
excentricitySquared = initializer.excentricitySquared.value;
- excentricity = sqrt(excentricitySquared);
+ excentricity = sqrt(excentricitySquared); //
DoubleDouble.sqrt() does not make any difference here.
inverse = new Inverse();
}
