JMSLTM Numerical Library 6.1

com.imsl.stat
Class Ranks

java.lang.Object
  extended by com.imsl.stat.Ranks

public class Ranks
extends Object

Compute the ranks, normal scores, or exponential scores for a vector of observations.

The class Ranks can be used to compute the ranks, normal scores, or exponential scores of the data in X. Ties in the data can be resolved in four different ways, as specified by member function setTieBreaker. The type of values returned can vary depending on the member function called:

GetRanks: Ordinary Ranks

For this member function, the values output are the ordinary ranks of the data in X. If X[i] has the smallest value among those in X and there is no other element in X with this value, then getRanks(i) = 1. If both X[i] and X[j] have the same smallest value, then

if TieBreaker = 0, Ranks[i] = getRanks([j] = 1.5
if TieBreaker = 1, Ranks[i] = Ranks[j] = 2.0
if TieBreaker = 2, Ranks[i] = Ranks[j] = 1.0
if TieBreaker = 3, Ranks[i] = 1.0 and Ranks[j] = 2.0
or Ranks[i] = 2.0 and Ranks[j] = 1.0.

When the ties are resolved by use of function setRandom, different results may occur when running the same program at different times unless the "seed" of the random number generator is set explicitly by use of Random method setSeed. Ordinarily, there is no need to call the routine to set the seed, even if there are ties in the data.

getBlomScores: Normal Scores, Blom Version

Normal scores are expected values, or approximations to the expected values, of order statistics from a normal distribution. The simplest approximations are obtained by evaluating the inverse cumulative normal distribution function, inverseNormal, at the ranks scaled into the open interval (0, 1). In the Blom version (see Blom 1958), the scaling transformation for the rank r_i(1 le r_i le n, where n is the sample size is (r_i - 3/8)/(n + 1/4). The Blom normal score corresponding to the observation with rank r_i is

Phi ^{ - 1} left( {frac{{r_i  - 3/8}}{{n + 1/4}}} 
  right)

where Phi(cdot) is the normal cumulative distribution function.

Adjustments for ties are made after the normal score transformation. That is, if X[i] equals X[j] (within fuzz) and their value is the k-th smallest in the data set, the Blom normal scores are determined for ranks of k and k + 1, and then these normal scores are averaged or selected in the manner specified by TieBreaker, which is set by the method setTieBreaker. (Whether the transformations are made first or ties are resolved first makes no difference except when averaging is done.)

getTukeyScores: Normal Scores, Tukey Version

In the Tukey version (see Tukey 1962), the scaling transformation for the rank r_i is (r_i - 1/3)/(n + 1/3). The Tukey normal score corresponding to the observation with rank r_i is

Phi ^{ - 1} left( {frac{{r_i  - 1/3}}{{n + 
  1/3}}} right)

Ties are handled in the same way as discussed above for the Blom normal scores.

getVanDerWaerdenScores: Normal Scores, Van der Waerden Version

In the Van der Waerden version (see Lehmann 1975, page 97), the scaling transformation for the rank r_i is r_i/(n + 1). The Van der Waerden normal score corresponding to the observation with rank r_i is

Phi ^{ - 1} left( {frac{{r_i }}{{n + 1}}} 
  right)

Ties are handled in the same way as discussed above for the Blom normal scores.

getNormalScores: Expected Value of Normal Order Statistics

The method getNormalScores returns the expected values of the normal order statistics. If the value in X[i] is the k-th smallest, then the value getNormalScores[i] is E(Z_k), where E(cdot) is the expectation operator and Z_k is the k-th order statistic in a sample of size NOBS from a standard normal distribution. Ties are handled in the same way as discussed above for the Blom normal scores.

getSavageScores: Savage Scores

The method getSavageScores returns the expected values of the exponential order statistics. These values are called Savage scores because of their use in a test discussed by Savage (1956) (see Lehman 1975). If the value in X[i] is the k-th smallest, then the i-th output value output is E(Y_k), where Y_k is the k-th order statistic in a sample of size n from a standard exponential distribution. The expected value of the k-th order statistic from an exponential sample of size n is

frac{1}{n} + frac{1}{{n - 1}} +  ldots + 
  frac{1}{{n - k + 1}}

Ties are handled in the same way as discussed above for the Blom normal scores.

See Also:
Example

Field Summary
static int TIE_AVERAGE
          In case of ties, use the average of the scores of the tied observations.
static int TIE_HIGHEST
          In case of ties, use the highest score in the group of ties.
static int TIE_LOWEST
          In case of ties, use the lowest score in the group of ties.
static int TIE_RANDOM
          In case of ties, use one of the group of ties chosen at random.
 
Constructor Summary
Ranks()
          Constructor for the Ranks class.
 
Method Summary
static double expectedNormalOrderStatistic(int i, int n)
          Returns the expected value of a normal order statistic.
 double[] getBlomScores(double[] x)
          Gets the Blom version of normal scores for each observation.
 double[] getNormalScores(double[] x)
          Gets the expected value of normal order statistics (for tied observations, the average of the expected normal scores).
 double[] getRanks(double[] x)
          Gets the rank for each observation.
 double[] getSavageScores(double[] x)
          Gets the Savage scores (the expected value of exponential order statistics).
 double[] getTukeyScores(double[] x)
          Gets the Tukey version of normal scores for each observation.
 double[] getVanDerWaerdenScores(double[] x)
          Gets the Van der Waerden version of normal scores for each observation.
 void setFuzz(double fuzz)
          Sets the fuzz factor used in determining ties.
 void setRandom(Random random)
          Sets the Random object.
 void setTieBreaker(int iTie)
          Sets the tie breaker for Ranks.
 
Methods inherited from class java.lang.Object
clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait
 

Field Detail

TIE_AVERAGE

public static final int TIE_AVERAGE
In case of ties, use the average of the scores of the tied observations.

See Also:
Constant Field Values

TIE_HIGHEST

public static final int TIE_HIGHEST
In case of ties, use the highest score in the group of ties.

See Also:
Constant Field Values

TIE_LOWEST

public static final int TIE_LOWEST
In case of ties, use the lowest score in the group of ties.

See Also:
Constant Field Values

TIE_RANDOM

public static final int TIE_RANDOM
In case of ties, use one of the group of ties chosen at random.

See Also:
Constant Field Values
Constructor Detail

Ranks

public Ranks()
Constructor for the Ranks class.

Method Detail

expectedNormalOrderStatistic

public static double expectedNormalOrderStatistic(int i,
                                                  int n)
Returns the expected value of a normal order statistic.

Parameters:
i - an int, the rank of the order statistic
n - an int, the sample size
Returns:
a double, the expected value of the i-th order statistic in a sample of size n from the standard normal distribution

getBlomScores

public double[] getBlomScores(double[] x)
Gets the Blom version of normal scores for each observation.

Parameters:
x - a double array which contains the observations to be ranked
Returns:
a double array which contains the Blom version of normal scores for each observation in x

getNormalScores

public double[] getNormalScores(double[] x)
Gets the expected value of normal order statistics (for tied observations, the average of the expected normal scores).

Parameters:
x - a double array which contains the observations
Returns:
a double array which contains the expected value of normal order statistics for the observations in x (for tied observations, the average of the expected normal scores)

getRanks

public double[] getRanks(double[] x)
Gets the rank for each observation.

Parameters:
x - a double array which contains the observations to be ranked
Returns:
a double array which contains the rank for each observation in x

getSavageScores

public double[] getSavageScores(double[] x)
Gets the Savage scores (the expected value of exponential order statistics).

Parameters:
x - a double array which contains the observations
Returns:
a double array which contains the Savage scores for the observations in x. (the expected value of exponential order statistics)

getTukeyScores

public double[] getTukeyScores(double[] x)
Gets the Tukey version of normal scores for each observation.

Parameters:
x - a double array which contains the observations to be ranked
Returns:
a double array which contains the Tukey version of normal scores for each observation in x

getVanDerWaerdenScores

public double[] getVanDerWaerdenScores(double[] x)
Gets the Van der Waerden version of normal scores for each observation.

Parameters:
x - a double array which contains the observations to be ranked
Returns:
a double array which contains the Van der Waerden version of normal scores for each observation in x

setFuzz

public void setFuzz(double fuzz)
Sets the fuzz factor used in determining ties.

Parameters:
fuzz - a double which represents the fuzz factor

setRandom

public void setRandom(Random random)
Sets the Random object.

Parameters:
random - a Random object used in breaking ties

setTieBreaker

public void setTieBreaker(int iTie)
Sets the tie breaker for Ranks.

Parameters:
iTie - an int which represents the tie breaker

JMSLTM Numerical Library 6.1

Copyright © 1970-2010 Visual Numerics, Inc.
Built July 30 2010.