uk.ac.starlink.ttools.plot2.data
Class FloatingCoord

java.lang.Object
  uk.ac.starlink.ttools.plot2.data.SingleCoord
      uk.ac.starlink.ttools.plot2.data.FloatingCoord

All Implemented Interfaces:

Coord

public class FloatingCoord
extends SingleCoord

Coord implementation for floating point values. This covers both single and double precision. Currently there is a factory method to generate an instance of this class, which uses single or double precision according to a global configuration parameter. Although in general double precision processing is important, for plotting purposes it's not usually going to make a visible difference, and the cached storage requirements are cut in half if you use single precision instead. Maybe see about different configuration options for this in the future.

Since:

4 Feb 2013

Author:

Mark Taylor

Field Summary

static FloatingCoord

WEIGHT_COORD Coordinate instance used for weighting values.

Method Summary

static FloatingCoord	createCoord(InputMeta meta, boolean isRequired) Factory method to return an instance of this class.
static FloatingCoord	createTimeCoord(InputMeta meta, boolean isRequired) Returns a new time coordinate.
Object	inputToStorage(Object[] userValues, uk.ac.starlink.table.DomainMapper[] mappers) Turns a quantity in the user view to a plotting view object.
double	readDoubleCoord(Tuple tuple, int icol) Reads a floating point value from an appropriate field in a given Tuple.

Methods inherited from class uk.ac.starlink.ttools.plot2.data.SingleCoord

getInput, getInputs, getStorageType, isRequired

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Field Detail

WEIGHT_COORD

public static FloatingCoord WEIGHT_COORD

Coordinate instance used for weighting values.

Method Detail

inputToStorage

public Object inputToStorage(Object[] userValues,
                             uk.ac.starlink.table.DomainMapper[] mappers)

Description copied from interface: Coord

Turns a quantity in the user view to a plotting view object. The return value is never null.

The supplied parameters both correspond (have the same length as) this object's Inputs array. For each Input, the corresponding element of the inputValues array gives the value obtained from the user-supplied data (matching Input.getValueClass()), and the corresponding element of the inputMappers array gives a DomainMapper object (consistent with Input.getDomain()). InputMappers may be null however, and in many cases, coordinates are not sensitive to domains, and for those cases implementations will ignore inputMappers.

Parameters:

userValues - per-input values

mappers - per-input domain mappers, each may be null

Returns:

object of the type corresponding to the result of Coord.getStorageType(); not null

readDoubleCoord

public double readDoubleCoord(Tuple tuple,
                              int icol)

Reads a floating point value from an appropriate field in a given Tuple.

Parameters:

tuple - tuple

icol - index of column in tuple corresponding to this Coord

Returns:

value of floating point field

createCoord

public static FloatingCoord createCoord(InputMeta meta,
                                        boolean isRequired)

Factory method to return an instance of this class. Implementation is currently determined by the PlotUtil.storeFullPrecision() method.

Parameters:

meta - input value metadata

isRequired - true if this coordinate is required for plotting

Returns:

instance

createTimeCoord

public static FloatingCoord createTimeCoord(InputMeta meta,
                                            boolean isRequired)

Returns a new time coordinate. This works in the TimeDomain, and the numeric value returned should normally be seconds since the Unix epoch (1 Jan 1970 midnight).

Parameters:

meta - input value metadata

isRequired - true if this coordinate is required for plotting

Returns:

instance