Module org.apache.sis.referencing
Package org.apache.sis.referencing.cs

Class CoordinateSystems

Object
Static
CoordinateSystems
public final class CoordinateSystems extends Static
Utility methods working on Coordinate­System objects and their axes. Those methods allow for example to estimate an angle between two axes or determining the change of axis directions and units between two coordinate systems.
Since:
0.4

  • Method Details

    • isGeodetic

      public static boolean isGeodetic(CoordinateSystem cs)
      Returns whether the given coordinate system can be associated to a Geodetic­CRS. This is true for instances of Ellipsoidal­CS, Cartesian­CS and Spherical­CS, and false for all other types of coordinate system.
      Parameters:
      cs - the coordinate system to test (can be null).
      Returns:
      whether the given coordinate system can be associated to a geodetic CRS.
      Since:
      1.3

    • parseAxisDirection

      public static AxisDirection parseAxisDirection(String name) throws IllegalArgumentException
      Returns an axis direction code from the given direction name. Names are case-insensitive. They may be:
      • Cardinal directions like "north" and "east".
      • Inter-cardinal directions "north-east" and "south-south-east", using either '-', '_' or spaces as separator between the cardinal points.
      • Directions from a pole like "South along 180 deg" and "South along 90° East", using either the "deg" or "°" symbol. Note that the meridian is not necessarily relative to Greenwich (see direction­Along­Meridian(…) for more information).
      Parameters:
      name - the direction name (e.g. "north", "north-east", etc.).
      Returns:
      the axis direction for the given name.
      Throws:
      Illegal­Argument­Exception - if the given name is not a known axis direction.

    • directionAlongMeridian

      public static AxisDirection directionAlongMeridian(AxisDirection baseDirection, double meridian)
      Returns an axis direction from a pole along a meridian. The given meridian is usually, but not necessarily, relative to the Greenwich meridian.

      Reference meridian

      The reference meridian depends on the context. It is usually the prime meridian of the geodetic datum of the geographic CRS instance that contains (through its coordinate system) the axes having those directions. This policy is consistent with WKT 2 specification §7.5.4(iv).

      Example

      direction­Along­Meridian(Axis­Direction​.SOUTH, -90) returns an axis direction for “South along 90°W”.
      Parameters:
      base­Direction - the base direction, which must be Axis­Direction​.NORTH or Axis­Direction​.SOUTH.
      meridian - the meridian in degrees, relative to a unspecified (usually Greenwich) prime meridian. Meridians in the East hemisphere are positive and meridians in the West hemisphere are negative.
      Returns:
      the axis direction along the given meridian.
      Since:
      0.6

    • isAlongMeridian

      public static boolean isAlongMeridian(AxisDirection direction)
      Returns true if the given axis direction seems to be a direction along a meridian.
      Examples
      Axis name Return value
      North along 90 deg East true
      South along 90 deg East true
      South false
      East false
      Note that true is not a guarantee that parse­Axis­Direction(String) will succeed. But it means that there is reasonable chances of success based on brief inspection of axis name.
      Parameters:
      direction - the direction to test. Can be null.
      Returns:
      if the given direction is non-null and seems to be a direction along a meridian.
      Since:
      1.2

    • angle

      public static Angle angle(AxisDirection source, AxisDirection target)
      Returns the arithmetic (counterclockwise) angle from the first axis direction to the second direction. This method returns a value between -180° and +180°, or null if no angle can be computed.

      Horizontal directions

      For any pair of compass directions which are not opposite directions, a positive angle denotes a right-handed system while a negative angle denotes a left-handed system. Examples:
      • The angle from EAST to NORTH is 90°
      • The angle from SOUTH to WEST is -90°
      • The angle from "North along 90° East" to "North along 0°" is 90°.
      In the case of directions like “South along 90°W”, the caller is responsible to make sure that the meridians are relative to the same prime meridian. This is the case if the axes are part of the same Coordinate­System instance.

      Horizontal and vertical directions

      By convention this method defines the angle from any compass direction to the UP vertical direction as 90°, and the angle of any compass direction to the DOWN vertical direction as -90°. The sign of those angles gives no indication about whether the coordinate system is right-handed or left-handed. Those angles are returned as instances of Elevation­Angle.

      All angles are approximations since this method does not take the Earth ellipsoidal or geoidal shape in account.

      Invariants

      For any non-null return value:
      • angle(A, A) = 0°
      • angle(A, opposite(A)) = ±180°
      • angle(A, B) = -angle(B, A)
      Parameters:
      source - the source axis direction.
      target - the target axis direction.
      Returns:
      the arithmetic angle (in degrees) of the rotation to apply on a line pointing toward the source direction in order to make it point toward the target direction, or null if this value cannot be computed.

    • swapAndScaleAxes

      public static Matrix swapAndScaleAxes(CoordinateSystem sourceCS, CoordinateSystem targetCS) throws IllegalArgumentException, IncommensurableException
      Returns an affine transform between two coordinate systems. Only units and axes order (e.g. transforming from (North, West) to (East, North) are taken in account by this method.

      Conditions

      The two coordinate systems must implement the same GeoAPI coordinate system interface. For example if source­CS is a Cartesian­CS, then target­CS must be a Cartesian­CS too.

      Example

      If coordinates in source­CS are (x,y) tuples in metres and coordinates in target­CS are (-y,x) tuples in centimetres, then the transformation can be performed as below: 
           ┌      ┐   ┌                ┐ ┌     ┐
     │-y(cm)│   │   0  -100    0 │ │ x(m)│
     │ x(cm)│ = │ 100     0    0 │ │ y(m)│
     │ 1    │   │   0     0    1 │ │ 1   │
     └      ┘   └                ┘ └     ┘
      Parameters:
      source­CS - the source coordinate system.
      target­CS - the target coordinate system.
      Returns:
      the conversion from source­CS to target­CS as an affine transform. Only axis direction and units are taken in account.
      Throws:
      Illegal­Argument­Exception - if the CS are not of the same type, or axes do not match.
      Incommensurable­Exception - if the units are not compatible, or the conversion is non-linear.
      See Also:

    • replaceAxes

      public static CoordinateSystem replaceAxes(CoordinateSystem cs, AxisFilter filter)
      Returns a coordinate system derived from the given one but with a modified list of axes. The axes may be filtered (excluding some axes), reordered or have their unit and direction modified.

      Example

      for replacing all angular units of a coordinate system to degrees (regardless what the original angular units were) while leaving other kinds of units unchanged, one can write: 
      CoordinateSystem cs = ...;
cs = CoordinateSystems.replaceAxes(cs, new AxisFilter() {
    @Override
    public Unit<?> getUnitReplacement(CoordinateSystemAxis axis, Unit<?> unit) {
        if (Units.isAngular(unit)) {
            unit = Units.DEGREE;
        }
        return unit;
    }
});

      Coordinate system normalization

      This method is often used together with swap­And­Scale­Axes(…) for normalizing the coordinate values given to a math transform. For example: 
      CoordinateSystem sourceCS = ...;
CoordinateSystem targetCS = ...;
Matrix step1 = swapAndScaleAxes(sourceCS, replaceAxes(sourceCS, AxisConvention.NORMALIZED));
Matrix step2 = ...; // some transform working on coordinates with standard axis order and unit.
Matrix step3 = swapAndScaleAxes(replaceAxes(targetCS, AxisConvention.NORMALIZED), targetCS);
      Parameters:
      cs - the coordinate system, or null.
      filter - the modifications to apply on coordinate system axes.
      Returns:
      the modified coordinate system as a new instance, or cs if the given coordinate system was null or does not need any change.
      Throws:
      Illegal­Argument­Exception - if the specified coordinate system cannot be filtered. It may be because the coordinate system would contain an illegal number of axes, or because an axis would have an unexpected direction or unexpected unit of measurement.
      Since:
      0.6
      See Also:

    • replaceLinearUnit

      public static CoordinateSystem replaceLinearUnit(CoordinateSystem cs, Unit<Length> newUnit)
      Returns a coordinate system derived from the given one but with all linear units replaced by the given unit. Non-linear units (e.g. angular or scale units) are left unchanged.

      This convenience method is equivalent to the following code:

      return CoordinateSystems.replaceAxes(cs, new AxisFilter() {
    @Override public Unit<?> getUnitReplacement(CoordinateSystemAxis axis, Unit<?> unit) {
        return Units.isLinear(unit) ? newUnit : unit;
    }
});
      Parameters:
      cs - the coordinate system in which to replace linear units, or null.
      new­Unit - the new linear unit.
      Returns:
      the modified coordinate system as a new instance, or null if the given cs was null, or cs if all linear units were already equal to the given one.
      Since:
      0.7
      See Also:

    • replaceAngularUnit

      public static CoordinateSystem replaceAngularUnit(CoordinateSystem cs, Unit<Angle> newUnit)
      Returns a coordinate system derived from the given one but with all angular units replaced by the given unit. Non-angular units (e.g. linear or scale units) are left unchanged.

      This convenience method is equivalent to the following code:

      return CoordinateSystems.replaceAxes(cs, new AxisFilter() {
    @Override public Unit<?> getUnitReplacement(CoordinateSystemAxis axis, Unit<?> unit) {
        return Units.isAngular(unit) ? newUnit : unit;
    }
});
      Parameters:
      cs - the coordinate system in which to replace angular units, or null.
      new­Unit - the new angular unit.
      Returns:
      the modified coordinate system as a new instance, or null if the given cs was null, or cs if all angular units were already equal to the given one.
      Since:
      0.7
      See Also:

    • getAxisDirections

      public static AxisDirection[] getAxisDirections(CoordinateSystem cs)
      Returns the axis directions for the specified coordinate system. This method guarantees that the returned array is non-null and does not contain any null direction.
      Parameters:
      cs - the coordinate system.
      Returns:
      the axis directions for the specified coordinate system.
      Throws:
      Null­Pointer­Exception - if cs is null, or one of its axes is null, or a value returned by Coordinate­System­Axis​.get­Direction() is null.
      Since:
      0.8

    • getShortName

      public static String getShortName(CoordinateSystemAxis axis, Locale locale)
      Returns a short (if possible) localized name for the given axis. This method replaces names such as "Geodetic latitude" or "Geocentric latitude" by a simple "Latitude" word. This method can be used for example in column or row headers when the context is known and the space is rare.
      Parameters:
      axis - the axis for which to get a short label.
      locale - desired locale for the label.
      Returns:
      a relatively short axis label, in the desired locale if possible.
      Since:
      1.3

    • getEpsgCode

      public static Integer getEpsgCode(Class<? extends CoordinateSystem> type, CoordinateSystemAxis... axes)
      Returns the EPSG code of a coordinate system using the units and directions of given axes. This method ignores axis metadata (names, abbreviation, identifiers, remarks, etc.). The axis minimum and maximum values are checked only if the range meaning is "wraparound". If no suitable coordinate system is known to Apache SIS, then this method returns null.

      Current implementation uses a hard-coded list of known coordinate systems; it does not yet scan the EPSG database (this may change in future Apache SIS version). The current list of known coordinate systems is given below.

      Known coordinate systems (CS)
      EPSG CS type Axis directions Horizontal unit
      6424 Ellipsoidal east north degree
      6422 Ellipsoidal north east degree
      6425 Ellipsoidal east north grads
      6403 Ellipsoidal north east grads
      6429 Ellipsoidal east north radian
      6428 Ellipsoidal north east radian
      6426 Ellipsoidal east north up degree
      6423 Ellipsoidal north east up degree
      6427 Ellipsoidal east north up grads
      6421 Ellipsoidal north east up grads
      6431 Ellipsoidal east north up radian
      6430 Ellipsoidal north east up radian
      4400 Cartesian east north metre
      4500 Cartesian north east metre
      4491 Cartesian west north metre
      4501 Cartesian north west metre
      6503 Cartesian west south metre
      6501 Cartesian south west metre
      1039 Cartesian east north foot
      1029 Cartesian north east foot
      4403 Cartesian east north Clarke’s foot
      4502 Cartesian north east Clarke’s foot
      4497 Cartesian east north US survey foot
      Parameters:
      type - the type of coordinate system for which an EPSG code is desired, as a GeoAPI interface.
      axes - axes for which a coordinate system EPSG code is desired.
      Returns:
      EPSG codes for a coordinate system using the given axes (ignoring metadata), or null if unknown to this method. Note that a null value does not mean that a more extensive search in the EPSG database would not find a matching coordinate system.
      Since:
      1.0
      See Also:

    • getEpsgCode

      public static Integer getEpsgCode(Unit<?> unit, AxisDirection... directions)
      Returns the EPSG code of a coordinate system using the given unit and axis directions. This convenience method performs the work documented in get­Epsg­Code(Class, Coordinate­System­Axis...), but requiring only a frequently used subset of information. If no suitable coordinate system is known to Apache SIS, then this method returns null.

      Current implementation uses a hard-coded list of known coordinate systems; it does not yet scan the EPSG database (this may change in future Apache SIS version). The current list of known coordinate systems is documented above.

      Parameters:
      unit - desired unit of measurement. For three-dimensional ellipsoidal coordinate system, this is the unit for the horizontal axes only; the vertical axis is in metres.
      directions - desired axis directions.
      Returns:
      EPSG codes for a coordinate system using the given axis directions and unit of measurement, or null if unknown to this method. Note that a null value does not mean that a more extensive search in the EPSG database would not find a matching coordinate system.
      Since:
      0.8
      See Also: