Module org.apache.sis.referencing

Package org.apache.sis.referencing.operation.transform

Class ContextualParameters

Object

Parameters

ContextualParameters

All Implemented Interfaces:

Serializable, Cloneable, General­Parameter­Value, Parameter­Value­Group

public class ContextualParameters
extends Parameters
implements Serializable

The parameters that describe a sequence of normalize → non-linear kernel → denormalize transforms as a whole. The normalize and denormalize steps must be affine transforms, while the non-linear kernel is arbitrary. This separation is useful for doing in the linear steps any unit conversions and axis swapping necessary for allowing the non-linear step to work in some predefined (implementation specific) coordinate system.

Contextual parameters are associated to the non-linear kernel step of the above-cited sequence. Since the parameter values of the non-linear kernel contains only normalized parameters (e.g. a map projection on an ellipsoid having a semi-major axis length of 1), Apache SIS needs contextual information for reconstructing the parameters of the complete transforms chain.

Usage in map projections

This object is used mostly for Apache SIS implementation of map projections, where the non-linear kernel is a normalized projection working on (longitude, latitude) coordinates in radians. The complete map projection (ignoring changes of axis order) is a chain of 3 transforms as shown below:

$$\left[\begin{array}{ccc}\pi /180& 0& -{\lambda }_0\cdot (\pi /180)\\ 0& \pi /180& 0\\ 0& 0& 1\end{array}\right]$$

→

Map projection on a normalized ellipsoid

→

$$\left[\begin{array}{ccc}a\cdot k_0& 0& \mathrm{FE}\\ 0& a\cdot k_0& \mathrm{FN}\\ 0& 0& 1\end{array}\right]$$

Contextual­Parameters is typically created and used as below:

1. A Math­Transform­Provider instantiates a non-linear transform working in a predefined coordinate system. Note that different providers may instantiate the same transform class but with different parameters. For example, both "Mercator (variant A)" and "Mercator (variant B)" operation methods instantiate the same Mercator projection class, but with different ways to represent the parameters.
2. The map projection constructor fetches all parameters that it needs from the user supplied Parameters, initializes the projection, then saves the parameter values that it actually used in a new Contextual­Parameters instance.
3. The map projection constructor may keep only the non-linear parameters for itself, and gives the linear parameters to the normalize­Geographic­Inputs(…) and Matrix­SIS.convert­After(…) methods, which will create the matrices show above. The projection constructor is free to apply additional operations on the two affine transforms (normalize / denormalize) before or after the above-cited methods have been invoked.
4. After all parameter values have been set and the normalize / denormalize matrices defined, the complete­Transform(…) method will mark this Contextual­Parameters object as unmodifiable and create the chain of transforms from (λ,φ) in angular degrees to (x,y) in metres. Note that conversions to other units and changes in axis order are not the purpose of this transforms chain – they are separated steps.

Serialization

Serialized instances of this class are not guaranteed to be compatible with future SIS versions. Serialization should be used only for short term storage or RMI between applications running the same SIS version.

Since:

0.6

See Also:

• Abstract­Math­Transform.get­Contextual­Parameters()
• Serialized Form

Nested Class Summary

Nested Classes

Modifier and Type	Class	Description
static enum	Contextual­Parameters.Matrix­Role	Whether a matrix is used for normalization or denormalization before or after a non-linear operation.

Constructor Summary

Constructors

Constructor	Description
Contextual­Parameters(Parameter­Descriptor­Group descriptor, int src­Dim, int tgt­Dim)	Creates a new group of parameters with the given descriptor.

Method Summary

Modifier and Type	Method	Description
Parameter­Value­Group	add­Group(String name)	Unsupported operation, since Contextual­Parameters groups do not contain sub-groups.
Contextual­Parameters	clone()	Returns a modifiable clone of this parameter value group.
Math­Transform	complete­Transform(Math­Transform­Factory factory, Math­Transform kernel)	Marks this Contextual­Parameter as unmodifiable and creates the normalize → kernel → denormalize transforms chain.
Matrix­SIS	denormalize­Geographic­Outputs(double λ0)	Appends a denormalization step after the non-linear kernel, which will convert input coordinates in the two first dimensions from radians to degrees.
boolean	equals(Object object)	Compares the given object with the parameters for equality.
final Parameter­Descriptor­Group	get­Descriptor()	Returns the parameters for the normalize → non-linear kernel → denormalize sequence as a whole.
final Matrix­SIS	get­Matrix(Contextual­Parameters.Matrix­Role role)	Returns the affine transforms to be applied before or after the non-linear kernel operation.
List<Parameter­Value­Group>	groups(String name)	Unsupported operation, since Contextual­Parameters groups do not contain sub-groups.
int	hash­Code()	Returns a hash code value for this object.
Matrix­SIS	normalize­Geographic­Inputs(double λ0)	Prepends a normalization step converting input coordinates in the two first dimensions from degrees to radians.
Parameter­Value<?>	parameter(String name)	Returns the parameter value of the given name.
List<General­Parameter­Value>	values()	Returns an unmodifiable list containing all parameters in this group.

Methods inherited from class Parameters

boolean­Value, cast, cast, cast­Or­Wrap, copy, double­Value, double­Value, double­Value­List, get­Descriptors, get­Mandatory­Value, get­Member­Name, get­Or­Create, get­Value, get­Value­Domain, int­Value, int­Value­List, is­Unmodifiable, print, string­Value, to­String, unmodifiable, unmodifiable

Methods inherited from class Object

finalize, get­Class, notify, notify­All, wait, wait, wait

Constructor Details

ContextualParameters

public ContextualParameters(ParameterDescriptorGroup descriptor,
 int srcDim,
 int tgtDim)

Creates a new group of parameters with the given descriptor. The method parameters shall describe the normalize → non-linear kernel → denormalize sequence as a whole. After construction, callers shall:

• Set the relevant parameter values by calls to parameter(…).set­Value(…).
• Modify the element values in normalization / denormalization affine transforms, optionally by calls to the convenience methods in this class.
• Get the complete transforms chain with a call complete­Transform(…)

See class javadoc for more information.

Parameters:

descriptor - the parameter descriptor.

src­Dim - number of source dimensions.

tgt­Dim - number of target dimensions.

Since:

1.0

Method Details

getDescriptor

public final ParameterDescriptorGroup getDescriptor()

Returns the parameters for the normalize → non-linear kernel → denormalize sequence as a whole. This is the parameter descriptor of the operation method given to the constructor.

The values for those parameters are given by the values() method. Those values may be used by the normalization / denormalization linear steps, by the non-linear kernel, or both.

Specified by:

get­Descriptor in interface General­Parameter­Value

Specified by:

get­Descriptor in interface Parameter­Value­Group

Returns:

the description of the parameters.

getMatrix

public final MatrixSIS getMatrix(ContextualParameters.MatrixRole role)

Returns the affine transforms to be applied before or after the non-linear kernel operation. Immediately after construction, those matrices are modifiable identity matrices. Callers can modify the matrix element values, typically by calls to the Matrix­SIS.convert­Before(…) method. Alternatively, the following methods can be invoked for applying some frequently used configurations:

• normalize­Geographic­Inputs(double)
• denormalize­Geographic­Outputs(double)

After the complete­Transform(…) method has been invoked, the matrices returned by this method are unmodifiable.

Application to map projections

After the construction of a non-linear kernel working on (longitude, latitude) coordinates in radians, the matrices returned by projection.get­Contextual­Parameters().get­Matrix(…) are initialized to the values shown below.

Initial matrix coefficients after Normalized­Projection construction

getMatrix(NORMALIZATION)	getMatrix(DENORMALIZATION)
$$\left[\begin{array}{ccc}\pi /180& 0& -{\lambda }_0\cdot (\pi /180)\\ 0& \pi /180& 0\\ 0& 0& 1\end{array}\right]$$	$$\left[\begin{array}{ccc}a\cdot k_0& 0& \mathrm{FE}\\ 0& a\cdot k_0& \mathrm{FN}\\ 0& 0& 1\end{array}\right]$$

Parameters:

role - NORMALIZATION for fetching the normalization transform to apply before the kernel, DENORMALIZATION for the denormalization transform to apply after the kernel, or INVERSE_* for the inverse of the above-cited matrices.

Returns:

the matrix for the requested normalization or denormalization affine transform.

Since:

0.7

normalizeGeographicInputs

public MatrixSIS normalizeGeographicInputs(double λ0)

Prepends a normalization step converting input coordinates in the two first dimensions from degrees to radians. The normalization can optionally subtract the given λ₀ value (in degrees) from the longitude.

Invoking this method is equivalent to concatenating the normalization matrix with the following matrix. This will have the effect of applying the conversion described above before any other operation:

$$\left[\begin{array}{ccc}\pi /180& 0& -{\lambda }_0\cdot (\pi /180)\\ 0& \pi /180& 0\\ 0& 0& 1\end{array}\right]$$

Parameters:

λ0 - longitude of the central meridian, in degrees.

Returns:

the normalization affine transform as a matrix. Callers can change that matrix directly if they want to apply additional normalization operations.

Throws:

Illegal­State­Exception - if this Contextual­Parameter has been made unmodifiable.

denormalizeGeographicOutputs

public MatrixSIS denormalizeGeographicOutputs(double λ0)

Appends a denormalization step after the non-linear kernel, which will convert input coordinates in the two first dimensions from radians to degrees. After this conversion, the denormalization can optionally add the given λ₀ value (in degrees) to the longitude.

Invoking this method is equivalent to concatenating the denormalization matrix with the following matrix. This will have the effect of applying the conversion described above after the non-linear kernel operation:

$$\left[\begin{array}{ccc}\mathrm{180}/\mathrm{\pi }& 0& {\lambda }_0\\ 0& \mathrm{180}/\mathrm{\pi }& 0\\ 0& 0& 1\end{array}\right]$$

Parameters:

λ0 - longitude of the central meridian, in degrees.

Returns:

the denormalization affine transform as a matrix. Callers can change that matrix directly if they want to apply additional denormalization operations.

Throws:

Illegal­State­Exception - if this Contextual­Parameter has been made unmodifiable.

completeTransform

public MathTransform completeTransform(MathTransformFactory factory,
 MathTransform kernel)
                                throws FactoryException

Marks this Contextual­Parameter as unmodifiable and creates the normalize → kernel → denormalize transforms chain. This method shall be invoked only after the (de)normalization matrices have been set to their final values.

The transforms chain created by this method does not include any step for changing axis order or for converting to units other than degrees or metres. Such steps, if desired, should be defined outside Contextual­Parameters. Efficient concatenation of those steps will happen "under the hood".

Parameters:

factory - the factory to use for creating math transform instances.

kernel - the non-linear kernel which expects "normalized" coordinates in a predefined coordinate system.

Returns:

the concatenation of normalize → the given kernel → denormalize transforms.

Throws:

Factory­Exception - if an error occurred while creating a math transform instance.

parameter

public ParameterValue<?> parameter(String name)
                            throws ParameterNotFoundException

Returns the parameter value of the given name. Before the call to complete­Transform(…), this method can be used for setting parameter values like below:

parameter("Scale factor").setValue(0.9996);   // Scale factor of Universal Transverse Mercator (UTM) projections.

After the call to complete­Transform(…), the returned parameters are read-only.

Specified by:

parameter in interface Parameter­Value­Group

Parameters:

name - the name of the parameter to search.

Returns:

the parameter value for the given name.

Throws:

Parameter­Not­Found­Exception - if there is no parameter of the given name.

values

public List<GeneralParameterValue> values()

Returns an unmodifiable list containing all parameters in this group. Callers should not attempt to modify the parameter values in this list.

Specified by:

values in interface Parameter­Value­Group

Returns:

all parameter values.

groups

public List<ParameterValueGroup> groups(String name)

Unsupported operation, since Contextual­Parameters groups do not contain sub-groups.

Specified by:

groups in interface Parameter­Value­Group

Parameters:

name - ignored.

Returns:

never returned.

addGroup

public ParameterValueGroup addGroup(String name)

Unsupported operation, since Contextual­Parameters groups do not contain sub-groups.

Specified by:

add­Group in interface Parameter­Value­Group

Parameters:

name - ignored.

Returns:

never returned.

clone

public ContextualParameters clone()

Returns a modifiable clone of this parameter value group.

Specified by:

clone in interface General­Parameter­Value

Specified by:

clone in interface Parameter­Value­Group

Overrides:

clone in class Parameters

Returns:

a clone of this parameter value group.

See Also:

• Parameters.copy(Parameter­Value­Group, Parameter­Value­Group)

hashCode

public int hashCode()

Returns a hash code value for this object. This value is implementation-dependent and may change in any future version.

Overrides:

hash­Code in class Object

equals

public boolean equals(Object object)

Compares the given object with the parameters for equality.

Overrides:

equals in class Object

Parameters:

object - the object to compare with the parameters.

Returns:

true if the given object is equal to this one.