Module org.apache.sis.feature
Package org.apache.sis.coverage

Class SampleDimension.Builder

Object
Builder
Enclosing class:
Sample­Dimension
public static class SampleDimension.Builder extends Object
A mutable builder for creating an immutable Sample­Dimension. The following properties can be set:
  • An optional name for the Sample­Dimension.
  • A single optional category for the background value.
  • An arbitrary number of qualitative categories.
  • An arbitrary number of quantitative categories.

A qualitative category is a range of sample values associated to a label. For example, 0 = no data, 1 = cloud, 2 = sea, 3 = land, etc. Missing values are also considered as a qualitative category and should be declared. If the missing value can be used as a background value for filling empty spaces in image resampling operations, then it should be declared using set­Background(…) method instead of add­Qualitative(…).

A quantitative category is a range of sample values associated to numbers with units of measurement. For example, 10 = 1.0°C, 11 = 1.1°C, 12 = 1.2°C, etc. A quantitative category has a transfer function (typically a scale factor and an offset) for converting sample values to values expressed in the unit of measurement.

Qualitative and quantitative categories can be mixed in the same Sample­Dimension, provided that their ranges do not overlap. After properties have been set, the sample dimension is created by invoking build().

Note for subclass implementations

Properties are ultimately set by the following methods: All other set­Name(…), add­Qualitative(…) and add­Quantitative(…) methods are convenience methods delegating to above-cited methods, thus providing single points that subclasses can override.
Since:
1.0

  • Constructor Details

    • Builder

      public Builder()
      Creates an initially empty builder for a sample dimension. Callers shall invoke at least one add­Foo(…) method before build().

  • Method Details

    • getName

      public GenericName getName()
      Returns the name specified by the last call to a set­Name(…) method. If set­Name(…) has not been invoked or if clear() has been invoked after, then this method returns null. In that case the build() method will default to the name of the first quantitative category if present, or "Untitled" (localized) otherwise.
      Returns:
      the name explicitly set by last call to set­Name(…), or null if none or cleared.
      Since:
      1.2
      See Also:

    • setName

      public SampleDimension.Builder setName(GenericName name)
      Sets an identification of the sample dimension. This is the value to be returned by Sample­Dimension​.get­Name(). If this method is invoked more than once, then the last specified name prevails (previous sample dimension names are discarded).
      Parameters:
      name - identification of the sample dimension.
      Returns:
      this, for method call chaining.

    • setName

      public SampleDimension.Builder setName(CharSequence name)
      Sets an identification of the sample dimension as a character sequence. This is a convenience method for creating a Generic­Name from the given characters.

      Default implementation

      This convenience method delegates to set­Name(Generic­Name).
      Parameters:
      name - identification of the sample dimension.
      Returns:
      this, for method call chaining.

    • setName

      public SampleDimension.Builder setName(int band)
      Sets an identification of the sample dimension as a band number. This method should be used only when no more descriptive name is available.

      Default implementation

      This convenience method delegates to set­Name(Generic­Name).
      Parameters:
      band - sequence identifier of the sample dimension to create.
      Returns:
      this, for method call chaining.

    • getBackground

      public Number getBackground()
      Returns the value specified by the last call to a set­Background(…) method. If set­Background(…) has not been invoked or if clear() has been invoked after, then this method returns null.
      Returns:
      the value set by last call to set­Background(…), or null if none or cleared.
      Since:
      1.2
      See Also:

    • setBackground

      public SampleDimension.Builder setBackground(Number sample)
      Sets the background value without creating a category (typically for RGB images). An image without category is interpreted as an image for visualization purposes only, without values that we can associate to a unit of measurement or a qualitative meaning. This set­Background(Number) method can be invoked when the caller nevertheless wants to declare a background value used for filling empty spaces (e.g. in image corners).
      Parameters:
      sample - the background value, or null if none.
      Returns:
      this, for method call chaining.
      Since:
      1.2

    • setBackground

      public SampleDimension.Builder setBackground(CharSequence name, Number sample)
      Adds a qualitative category and marks that category as the background value. This is the value to be returned by Sample­Dimension​.get­Background(). If this method is invoked more than once, then the last specified value prevails (previous values become ordinary qualitative categories).
      Parameters:
      name - the category name as a String or International­String object, or null for a default "fill value" (localized) name.
      sample - the background value.
      Returns:
      this, for method call chaining.

    • addQualitative

      public SampleDimension.Builder addQualitative(CharSequence name, boolean sample)
      Adds a qualitative category for samples of the given boolean value. The true value is represented by 1 and the false value is represented by 0.

      Usage note

      The set­Background(Char­Sequence, Number) method should be used instead of this method when the aim is to define a default "no data" category to use when the missing value cannot be categorized more precisely (cloud, instrument error, etc).

      Default implementation

      This convenience method delegates to add­Qualitative(Char­Sequence, Number­Range).
      Parameters:
      name - the category name as a String or International­String object, or null for a default "no data" name.
      sample - the sample value as a boolean.
      Returns:
      this, for method call chaining.

    • addQualitative

      public SampleDimension.Builder addQualitative(CharSequence name, byte sample)
      Adds a qualitative category for samples of the given tiny (8 bits) integer value. The argument is treated as a signed integer (0x80 to 0x7f).

      Usage note

      The set­Background(Char­Sequence, Number) method should be used instead of this method when the aim is to define a default "no data" category to use when the missing value cannot be categorized more precisely (cloud, instrument error, etc).

      Default implementation

      This convenience method delegates to add­Qualitative(Char­Sequence, Number­Range).
      Parameters:
      name - the category name as a String or International­String object, or null for a default "no data" name.
      sample - the sample value as an integer.
      Returns:
      this, for method call chaining.

    • addQualitative

      public SampleDimension.Builder addQualitative(CharSequence name, short sample)
      Adds a qualitative category for samples of the given short (16 bits) integer value. The argument is treated as a signed integer (-32768 to 32767).

      Usage note

      The set­Background(Char­Sequence, Number) method should be used instead of this method when the aim is to define a default "no data" category to use when the missing value cannot be categorized more precisely (cloud, instrument error, etc).

      Default implementation

      This convenience method delegates to add­Qualitative(Char­Sequence, Number­Range).
      Parameters:
      name - the category name as a String or International­String object, or null for a default "no data" name.
      sample - the sample value as an integer.
      Returns:
      this, for method call chaining.

    • addQualitative

      public SampleDimension.Builder addQualitative(CharSequence name, int sample)
      Adds a qualitative category for samples of the given integer value. The argument is treated as a signed integer (-2147483648 to 2147483647).

      Usage note

      The set­Background(Char­Sequence, Number) method should be used instead of this method when the aim is to define a default "no data" category to use when the missing value cannot be categorized more precisely (cloud, instrument error, etc).

      Default implementation

      This convenience method delegates to add­Qualitative(Char­Sequence, Number­Range).
      Parameters:
      name - the category name as a String or International­String object, or null for a default "no data" name.
      sample - the sample value as an integer.
      Returns:
      this, for method call chaining.

    • addQualitative

      public SampleDimension.Builder addQualitative(CharSequence name, float sample)
      Adds a qualitative category for samples of the given floating-point value.

      Usage note

      The set­Background(Char­Sequence, Number) method should be used instead of this method when the aim is to define a default "no data" category to use when the missing value cannot be categorized more precisely (cloud, instrument error, etc).

      Default implementation

      This convenience method delegates to add­Qualitative(Char­Sequence, Number­Range).
      Parameters:
      name - the category name as a String or International­String object, or null for a default "no data" name.
      sample - the sample value as a real number or a NaN value.
      Returns:
      this, for method call chaining.

    • addQualitative

      public SampleDimension.Builder addQualitative(CharSequence name, double sample)
      Adds a qualitative category for samples of the given double precision floating-point value.

      Usage note

      The set­Background(Char­Sequence, Number) method should be used instead of this method when the aim is to define a default "no data" category to use when the missing value cannot be categorized more precisely (cloud, instrument error, etc).

      Default implementation

      This convenience method delegates to add­Qualitative(Char­Sequence, Number­Range).
      Parameters:
      name - the category name as a String or International­String object, or null for a default "no data" name.
      sample - the sample value as a real number or a NaN value.
      Returns:
      this, for method call chaining.

    • addQualitative

      public SampleDimension.Builder addQualitative(CharSequence name, Number minimum, Number maximum)
      Adds a qualitative category for samples in the given range of values.

      Usage note

      The set­Background(Char­Sequence, Number) method should be used instead of this method when the aim is to define a default "no data" category to use when the missing value cannot be categorized more precisely (cloud, instrument error, etc).

      Default implementation

      This convenience method delegates to add­Qualitative(Char­Sequence, Number­Range).
      Parameters:
      name - the category name as a String or International­String object, or null for a default "no data" name.
      minimum - the minimum sample value, inclusive.
      maximum - the maximum sample value, inclusive.
      Returns:
      this, for method call chaining.
      Throws:
      Illegal­Argument­Exception - if the range is empty.

    • addQualitative

      public SampleDimension.Builder addQualitative(CharSequence name, NumberRange<?> samples)
      Adds a qualitative category for all samples in the specified range of values. This is the most generic method for adding a qualitative category. All other add­Qualitative(name, …) methods are convenience methods delegating their work to this method.

      Usage note

      The set­Background(Char­Sequence, Number) method should be used instead of this method when the aim is to define a default "no data" category to use when the missing value cannot be categorized more precisely (cloud, instrument error, etc).
      Parameters:
      name - the category name as a String or International­String object, or null for a default "no data" name.
      samples - the minimum and maximum sample values in the category.
      Returns:
      this, for method call chaining.
      Throws:
      Illegal­Argument­Exception - if the given range is empty.

    • mapQualitative

      public SampleDimension.Builder mapQualitative(CharSequence name, Number sample, float converted)
      Adds a qualitative category for the given sample value mapped to the specified converted NaN value. The given converted value must be a NaN value (not necessarily the Float​.Na­N canonical value) and that value shall not have been used by another category.

      Default implementation

      This convenience method delegates to map­Qualitative(Char­Sequence, Number­Range, float).
      Parameters:
      name - the category name as a String or International­String object, or null for a default "no data" name.
      sample - the sample value as a real or integer number.
      converted - the converted value to map to the given sample value. Float​.is­Na­N(converted) must be true.
      Returns:
      this, for method call chaining.
      Throws:
      Illegal­Argument­Exception - if converted is not a NaN value.
      Since:
      1.1
      See Also:

    • mapQualitative

      public SampleDimension.Builder mapQualitative(CharSequence name, NumberRange<?> samples, float converted)
      Adds a qualitative category for the given samples values mapped to the specified converted NaN value. The given converted value must be a NaN value (not necessarily the Float​.Na­N canonical value) and that value shall not have been used by another category.

      The add­Qualitative(…) methods select automatically a converted value. The map­Qualitative(…) methods are for the less common case where caller needs to control which converted NaN value is mapped to the samples values.

      Parameters:
      name - the category name as a String or International­String object, or null for a default "no data" name.
      samples - the minimum and maximum sample values in the category.
      converted - the converted value to map to the given sample values. Float​.is­Na­N(converted) must be true.
      Returns:
      this, for method call chaining.
      Throws:
      Illegal­Argument­Exception - if converted is not a NaN value.
      Since:
      1.1
      See Also:

    • addQuantitative

      public SampleDimension.Builder addQuantitative(CharSequence name, NumberRange<?> samples, NumberRange<?> converted)
      Constructs a quantitative category mapping samples to real values in the specified range. Sample values in the samples range will be mapped to real values in the converted range through a linear equation of the form:
      measure = sample × scale + offset
      where scale and offset coefficients are computed from the ranges supplied in arguments. The units of measurement will be taken from the converted range if it is an instance of Measurement­Range.

      Warning: this method is provided for convenience when the scale and offset factors are not explicitly specified. If those factor are available, then the other add­Quantitative(name, samples, …) methods are more reliable.

      Default implementation

      This convenience method delegates to add­Quantitative(Char­Sequence, Number­Range, Math­Transform1D, Unit).
      Parameters:
      name - the category name as a String or International­String object, or null for a default "data" name.
      samples - the minimum and maximum sample values in the category. Element class is usually Integer, but Float and Double values are accepted as well.
      converted - the range of real values for this category, as an instance of Measurement­Range if those values are associated to an unit of measurement.
      Returns:
      this, for method call chaining.
      Throws:
      Class­Cast­Exception - if the range element class is not a Number subclass.
      Illegal­Argument­Exception - if the range is invalid.

    • addQuantitative

      public SampleDimension.Builder addQuantitative(CharSequence name, float minimum, float maximum, Unit<?> units)
      Adds a quantitative category for values ranging from minimum to maximum inclusive in the given units of measurement. The transfer function is set to identity.

      Default implementation

      This convenience method delegates to add­Quantitative(Char­Sequence, Number­Range, Math­Transform1D, Unit).
      Parameters:
      name - the category name as a String or International­String object, or null for a default "data" name.
      minimum - the minimum value (inclusive) in the given units.
      maximum - the maximum value (inclusive) in the given units.
      units - the units of measurement, or null if unknown or not applicable.
      Returns:
      this, for method call chaining.
      Throws:
      Illegal­Argument­Exception - if a value is NaN or if minimum is greater than maximum.

    • addQuantitative

      public SampleDimension.Builder addQuantitative(CharSequence name, double minimum, double maximum, Unit<?> units)
      Adds a quantitative category for values ranging from minimum to maximum inclusive in the given units of measurement. The transfer function is set to identity.

      Default implementation

      This convenience method delegates to add­Quantitative(Char­Sequence, Number­Range, Math­Transform1D, Unit).
      Parameters:
      name - the category name as a String or International­String object, or null for a default "data" name.
      minimum - the minimum value (inclusive) in the given units.
      maximum - the maximum value (inclusive) in the given units.
      units - the units of measurement, or null if unknown or not applicable.
      Returns:
      this, for method call chaining.
      Throws:
      Illegal­Argument­Exception - if a value is NaN or if minimum is greater than maximum.

    • addQuantitative

      public SampleDimension.Builder addQuantitative(CharSequence name, int lower, int upper, double scale, double offset, Unit<?> units)
      Adds a quantitative category for sample values ranging from lower inclusive to upper exclusive. Sample values are converted into real values using the following linear equation:
      measure = sample × scale + offset
      Results of above conversion are measurements in the units specified by the units argument.

      Default implementation

      This convenience method delegates to add­Quantitative(Char­Sequence, Number­Range, Math­Transform1D, Unit).
      Parameters:
      name - the category name as a String or International­String object. or null for a default "data" name.
      lower - the lower sample value, inclusive.
      upper - the upper sample value, exclusive.
      scale - the scale value which is multiplied to sample values for the category. Must be different than zero.
      offset - the offset value to add to sample values for this category.
      units - the units of measurement of values after conversion by the scale factor and offset, or null if unknown or not applicable.
      Returns:
      this, for method call chaining.
      Throws:
      Illegal­Argument­Exception - if lower is not smaller than upper, or if scale or offset are not real numbers, or if scale is zero.

    • addQuantitative

      public SampleDimension.Builder addQuantitative(CharSequence name, NumberRange<?> samples, MathTransform1D toUnits, Unit<?> units)
      Constructs a quantitative category for all samples in the specified range of values. Sample values (usually integers) will be converted into real values (usually floating-point numbers) through the to­Units transform. Results of that conversion are measurements in the units specified by the units argument.

      This is the most generic method for adding a quantitative category. All other add­Quantitative(name, …) methods are convenience methods delegating their work to this method.

      Parameters:
      name - the category name as a String or International­String object, or null for a default "data" name.
      samples - the minimum and maximum sample values in the category. Element class is usually Integer, but Float and Double types are accepted as well.
      to­Units - the transfer function from sample values to real values in the specified units.
      units - the units of measurement of values after conversion by the transfer function, or null if unknown or not applicable.
      Returns:
      this, for method call chaining.
      Throws:
      Class­Cast­Exception - if the range element class is not a Number subclass.
      Illegal­Argument­Exception - if the range is invalid.
      See Also:

    • categories

      public List<Category> categories()
      Returns the list of categories added so far. The returned list does not support the add operation, but supports the remove operation.
      Returns:
      the current category list, read-only except for the remove operation.

    • build

      public SampleDimension build()
      Creates a new sample with the properties defined to this builder.
      Returns:
      the sample dimension.
      Throws:
      Illegal­Sample­Dimension­Exception - if there is overlapping ranges of sample values or other problems that prevent the construction of sample dimensions.

    • clear

      public void clear()
      Reset this builder to the same state than after construction. The sample dimension name, background values and all categories are discarded. This method can be invoked when the same builder is reused for creating more than one sample dimension.