Package net.imglib2.histogram

Class Histogram1d<T>

java.lang.Object
net.imglib2.histogram.Histogram1d<T>
All Implemented Interfaces:
Iterable<LongType>, Dimensions, EuclideanSpace, Img<LongType>, Interval, IterableInterval<LongType>, IterableRealInterval<LongType>, RandomAccessible<LongType>, RandomAccessibleInterval<LongType>, RealInterval
public class Histogram1d<T> extends Object implements Img<LongType>
A Histogram1d is a histogram that tracks up to four kinds of values: 1) values in the center of the distribution 2) values to the left of the center of the distribution (lower tail) 3) values to the right of the center of the distribution (upper tail) 4) values outside the other areas

Note: the last three classifications may not be present depending upon the makeup of the input data.

Author:
Barry DeZonia

  • Constructor Details

    • Histogram1d

      public Histogram1d(BinMapper1d<T> mapper)
      Construct a histogram from a bin mapping algorithm. Use countData() to populate it.
      Parameters:
      mapper - The algorithm used to map values to bins

    • Histogram1d

      public Histogram1d(Histogram1d<T> other)
      Construct a histogram whose bin mappings match another histogram. After this construction the histogram bins are unpopulated.
      Parameters:
      other - The histogram to copy.

    • Histogram1d

      public Histogram1d(Iterable<T> data, BinMapper1d<T> mapper)
      Construct a histogram from an iterable set of data and a bin mapping algorithm.
      Parameters:
      data - The iterable set of values to calculate upon
      mapper - The algorithm used to map values to bins

  • Method Details

    • firstDataValue

      public T firstDataValue()
      Returns the first data value of the input iteration.

    • hasTails

      public boolean hasTails()
      Returns true if the histogram has tail bins at both ends which count extreme values.

    • lowerTailCount

      public long lowerTailCount()
      Returns the frequency count of values in the lower tail bin (if any).

    • upperTailCount

      public long upperTailCount()
      Returns the frequency count of values in the upper tail bin (if any).

    • valueCount

      public long valueCount()
      Returns the frequency count of all values in the middle of the distribution.

    • distributionCount

      public long distributionCount()
      Returns the frequency count of all values in the distribution: lower tail + middle + upper tail. Does not include ignored values.

    • ignoredCount

      public long ignoredCount()
      Returns the frequency count of values that were ignored because they could not be mapped to any bin.

    • totalCount

      public long totalCount()
      Returns the total count of all values observed; both within and without the entire distribution. Thus it includes ignored values. One should decide carefully between using distributionCount() and totalCount().

    • frequency

      public long frequency(T value)
      Returns the frequency count of values within a bin using a representative value. Note that multiple values can be mapped to one bin so this is NOT the frequency count of this exact value in the distribution.
      Parameters:
      value - A representative value of interest

    • frequency

      public long frequency(long binPos)
      Returns the frequency count of the values within a bin.

    • relativeFrequency

      public double relativeFrequency(T value, boolean includeTails)
      Returns the relative frequency of values within a bin using a representative value. Note that multiple values can be mapped to one bin so this is NOT the relative frequency of this exact value in the distribution.

      This calculation is of the number of values in the bin divided by either the number of values in the distribution or the number of values in the center of the distribution (tails ignored).

      One can devise other ways to count relative frequencies that consider ignored values also. If needed one can use the various count methods and frequency methods to calculate any relative frequency desired.

      Parameters:
      value - A representative value of interest
      includeTails - Flag for determining whether to include tails in calculation.

    • relativeFrequency

      public double relativeFrequency(long binPos, boolean includeTails)
      Returns the relative frequency of values within a bin.

      This calculation is of the number of values in the bin divided by either the number of values in the distribution or the number of values in the center of the distribution (tails ignored).

      One can devise other ways to count relative frequencies that consider ignored values also. If needed one can use the various count methods and frequency methods to calculate any relative frequency desired.

      Parameters:
      binPos - The position of the bin of interest
      includeTails - Flag for determining whether to include tails in calculation.

    • getBinCount

      public long getBinCount()
      Returns the number of bins contained in the histogram.

    • map

      public long map(T value)
      Returns a bin position by mapping from a representative value.

    • getCenterValue

      public void getCenterValue(long binPos, T value)
      Gets the value associated with the center of a bin.
      Parameters:
      binPos - The bin number of interest
      value - The output to fill with the center value

    • getLowerBound

      public void getLowerBound(long binPos, T value)
      Gets the value associated with the left edge of a bin.
      Parameters:
      binPos - The bin number of interest
      value - The output to fill with the left edge value

    • getUpperBound

      public void getUpperBound(long binPos, T value)
      Gets the value associated with the right edge of the bin.
      Parameters:
      binPos - The bin number of interest
      value - The output to fill with the right edge value

    • includesUpperBound

      public boolean includesUpperBound(long binPos)
      Returns true if the given bin interval is closed on the right
      Parameters:
      binPos - The bin number of the interval of interest

    • includesLowerBound

      public boolean includesLowerBound(long binPos)
      Returns true if the given bin interval is closed on the left
      Parameters:
      binPos - The bin number of the interval of interest

    • isInLowerTail

      public boolean isInLowerTail(T value)
      Returns true if a given value is mapped to the lower tail of the distribution.
      Parameters:
      value - The value to determine the location of

    • isInUpperTail

      public boolean isInUpperTail(T value)
      Returns true if a given value is mapped to the upper tail of the distribution.
      Parameters:
      value - The value to determine the location of

    • isInMiddle

      public boolean isInMiddle(T value)
      Returns true if a given value is mapped to the middle of the distribution.
      Parameters:
      value - The value to determine the location of

    • isOutside

      public boolean isOutside(T value)
      Returns true if a given value is outside the distribution.
      Parameters:
      value - The value to determine the location of

    • dfd

      public DiscreteFrequencyDistribution dfd()
      Get the discrete frequency distribution associated with this histogram.

    • countData

      public void countData(Iterable<T> data)
      Counts the data contained in the given data source using the underlying bin distribution.
      Parameters:
      data - The total data to count

    • addData

      public void addData(Iterable<T> data)
      Counts additional data contained in a given iterable collection. One can use this to update an existing histogram with a subset of values.
      Parameters:
      data - The new data to count

    • subtractData

      public void subtractData(Iterable<T> data)
      Uncounts some original data contained in a given iterable collection. One can use this to update an existing histogram with a subset of values.
      Parameters:
      data - The old data to uncount

    • increment

      public void increment(long binPos)
      Directly increment a bin by position.
      Parameters:
      binPos - The 1-d index of the bin

    • decrement

      public void decrement(long binPos)
      Directly decrement a bin by position.
      Parameters:
      binPos - The 1-d index of the bin

    • increment

      public void increment(T value)
      Directly increment a bin by value.
      Parameters:
      value - The value to map to a bin position

    • decrement

      public void decrement(T value)
      Directly decrement a bin by value,
      Parameters:
      value - The value to map to a bin position

    • resetCounters

      public void resetCounters()
      Resets all data counts to 0.

    • toLongArray

      public long[] toLongArray()
      Returns a bare long[] histogram with the same bin counts as this histogram.

    • numDimensions

      public int numDimensions()
      Return the number of dimensions of the frequency distribution of this histogram.
      Specified by:
      numDimensions in interface EuclideanSpace

    • dimension

      public long dimension(int d)
      Return the size of the given dimension of the frequency distribution of this histogram.
      Specified by:
      dimension in interface Dimensions
      Parameters:
      d -

    • dimensions

      public void dimensions(long[] dims)
      Fill the provided long[] with the sizes of all dimensions of the frequency distribution of this histogram.
      Specified by:
      dimensions in interface Dimensions
      Parameters:
      dims -

    • randomAccess

      public RandomAccess<LongType> randomAccess()
      Description copied from interface: RandomAccessible
      Create a random access sampler for integer coordinates.

      The returned random access covers as much of the domain as possible.

      Please note: RandomAccessibleIntervals have a finite domain (their Interval), so RandomAccessible.randomAccess() is only guaranteed to cover this finite domain. This may lead to unexpected results when using Views. In the following code 
       RandomAccessible<T> extended = Views.extendBorder( img )
 RandomAccessibleInterval<T> cropped = Views.interval( extended, img );
 RandomAccess<T> a1 = extended.randomAccess();
 RandomAccess<T> a2 = cropped.randomAccess();
      The access a1 on the extended image is valid everywhere. However, somewhat counter-intuitively, the access a2 on the extended and cropped image is only valid on the interval img to which the extended image was cropped. The access is only required to cover this interval, because it is the domain of the cropped image. Views attempts to provide the fastest possible access that meets this requirement, and will therefore strip the extension. To deal with this, if you know that you need to access pixels outside the domain of the RandomAccessibleInterval, and you know that the RandomAccessibleInterval is actually defined beyond its interval boundaries, then use the RandomAccessible.randomAccess(Interval) variant and specify which interval you actually want to access. In the above example, 
       RandomAccess<T> a2 = cropped.randomAccess( Intervals.expand( img, 10 ) );
      will provide the extended access as expected.
      Specified by:
      randomAccess in interface RandomAccessible<T>
      Returns:
      random access sampler

    • randomAccess

      public RandomAccess<LongType> randomAccess(Interval interval)
      Description copied from interface: RandomAccessible
      Create a random access sampler for integer coordinates.

      The returned random access is intended to be used in the specified interval only. Thus, the RandomAccessible may provide optimized versions. If the interval is completely contained in the domain, the random access is guaranteed to provide the same values as that obtained by RandomAccessible.randomAccess() within the interval.

      Specified by:
      randomAccess in interface RandomAccessible<T>
      Parameters:
      interval - in which interval you intend to use the random access.
      Returns:
      random access sampler

    • min

      public long min()

    • min

      public long min(int d)
      Description copied from interface: Interval
      Get the minimum in dimension d.
      Specified by:
      min in interface Interval
      Parameters:
      d - dimension
      Returns:
      minimum in dimension d.

    • min

      public void min(long[] min)
      Description copied from interface: Interval
      Write the minimum of each dimension into long[].
      Specified by:
      min in interface Interval
      Parameters:
      min -

    • min

      public void min(Positionable min)
      Description copied from interface: Interval
      Sets a Positionable to the minimum of this Interval
      Specified by:
      min in interface Interval
      Parameters:
      min -

    • max

      public long max()

    • max

      public long max(int d)
      Description copied from interface: Interval
      Get the maximum in dimension d.
      Specified by:
      max in interface Interval
      Parameters:
      d - dimension
      Returns:
      maximum in dimension d.

    • max

      public void max(long[] max)
      Description copied from interface: Interval
      Write the maximum of each dimension into long[].
      Specified by:
      max in interface Interval
      Parameters:
      max -

    • max

      public void max(Positionable max)
      Description copied from interface: Interval
      Sets a Positionable to the maximum of this Interval
      Specified by:
      max in interface Interval
      Parameters:
      max -

    • realMin

      public double realMin()

    • realMin

      public double realMin(int d)
      Description copied from interface: RealInterval
      Get the minimum in dimension d.
      Specified by:
      realMin in interface RealInterval
      Parameters:
      d - dimension
      Returns:
      minimum in dimension d.

    • realMin

      public void realMin(double[] min)
      Description copied from interface: RealInterval
      Write the minimum of each dimension into double[].
      Specified by:
      realMin in interface RealInterval
      Parameters:
      min -

    • realMin

      public void realMin(RealPositionable min)
      Description copied from interface: RealInterval
      Sets a RealPositionable to the minimum of this Interval
      Specified by:
      realMin in interface RealInterval
      Parameters:
      min -

    • realMax

      public double realMax()

    • realMax

      public double realMax(int d)
      Description copied from interface: RealInterval
      Get the maximum in dimension d.
      Specified by:
      realMax in interface RealInterval
      Parameters:
      d - dimension
      Returns:
      maximum in dimension d.

    • realMax

      public void realMax(double[] max)
      Description copied from interface: RealInterval
      Write the maximum of each dimension into double[].
      Specified by:
      realMax in interface RealInterval
      Parameters:
      max -

    • realMax

      public void realMax(RealPositionable max)
      Description copied from interface: RealInterval
      Sets a RealPositionable to the maximum of this Interval
      Specified by:
      realMax in interface RealInterval
      Parameters:
      max -

    • cursor

      public Cursor<LongType> cursor()
      Description copied from interface: IterableRealInterval

      Returns a RealCursor that iterates with optimal speed without calculating the location at each iteration step. Localization is performed on demand.

      Use this where localization is required rarely/ not for each iteration.

      Specified by:
      cursor in interface IterableInterval<T>
      Specified by:
      cursor in interface IterableRealInterval<T>
      Returns:
      fast iterating iterator

    • localizingCursor

      public Cursor<LongType> localizingCursor()
      Description copied from interface: IterableRealInterval

      Returns a RealLocalizable Iterator that calculates its location at each iteration step. That is, localization is performed with optimal speed.

      Use this where localization is required often/ for each iteration.

      Specified by:
      localizingCursor in interface IterableInterval<T>
      Specified by:
      localizingCursor in interface IterableRealInterval<T>
      Returns:
      fast localizing iterator

    • size

      public long size()
      Description copied from interface: IterableRealInterval

      Returns the number of elements in this Function.

      Specified by:
      size in interface IterableRealInterval<T>
      Returns:
      number of elements

    • firstElement

      public LongType firstElement()
      Description copied from interface: IterableRealInterval
      Get the first element of this IterableRealInterval. This is a shortcut for cursor().next(). This can be used to create a new variable of type T using firstElement().createVariable(), which is useful in generic methods to store temporary results, e.g., a running sum over pixels in the IterableRealInterval.
      Specified by:
      firstElement in interface IterableRealInterval<T>
      Returns:
      the first element in iteration order.

    • iterationOrder

      public Object iterationOrder()
      Description copied from interface: IterableRealInterval
      Returns the iteration order of this IterableRealInterval. If the returned object equals (Object.equals(Object)) the iteration order of another IterableRealInterval f then they can be copied by synchronous iteration. That is, having an Iterator on this and another Iterator on f, moving both in synchrony will point both of them to corresponding locations in their source domain. In other words, this and f have the same iteration order and means and the same number of elements.
      Specified by:
      iterationOrder in interface IterableRealInterval<T>
      Returns:
      the iteration order of this IterableRealInterval.
      See Also:

    • iterator

      public Iterator<LongType> iterator()
      Specified by:
      iterator in interface Iterable<T>

    • factory

      public ImgFactory<LongType> factory()
      Description copied from interface: Img
      Get a ImgFactory that creates Imgs of the same kind as this one. This is useful to create Imgs for temporary storage in generic methods where the specific Img type is unknown. Note, that the factory can be used even if all references to this Img have been invalidated.
      Specified by:
      factory in interface Img<T>
      Returns:
      a factory for Imgs of the same kind as this one.

    • copy

      public Histogram1d<T> copy()
      Specified by:
      copy in interface Img<T>
      Returns:
      - A copy of the current Img instance, all pixels are duplicated