org.fhcrc.cpl.toolbox.statistics
Class BasicStatistics

java.lang.Object
  org.fhcrc.cpl.toolbox.statistics.BasicStatistics

public class BasicStatistics
extends java.lang.Object

A class with generic methods for basic stats. Add methods as needed. A lot of things are written twice, once for floats and once for doubles. That's annoying. But it would be wasteful of space and time to do conversions

Constructor Summary

BasicStatistics()

Method Summary

static double	calcNormalCumDensity(double mu, double sigma, double x) Calculate the cumulative density of the normal distribution function with parameters mu and sigma, at point x
static double	calcNormalDistDensity(double mu, double sigma, double x) Calculate the density of the normal distribution function with parameters mu and sigma, at point x
static double	calcStandardNormalCumDensity(double z) Calculate the cumulative density of the standard normal distribution at point x
static double	coefficientOfVariation(double[] input)
static double	coefficientOfVariation(java.util.List<? extends java.lang.Number> input)
static double	correlationCoefficient(double[] xvalues, double[] yvalues) covariance(xy) / (sx * sy)
static float	correlationCoefficient(float[] xvalues, float[] yvalues) covariance(xy) / (sx * sy)
static double	correlationCoefficient(java.util.List<? extends java.lang.Number> xvaluesList, java.util.List<? extends java.lang.Number> yvaluesList)
static double	covariance(double[] xvalues, double[] yvalues) Covariance = (sum(x*y) - ((sumx)(sumy)/n)) / (n-1)
static float	covariance(float[] xvalues, float[] yvalues) Covariance = (sum(x*y) - ((sumx)(sumy)/n)) / (n-1)
static double	erf(double z) Error function for standard normal distribution
static double	geometricMean(double[] values) Geometric mean is defined as ((x1)(x2)(x3)...(xk)) ^ 1/k
static double	geometricMean(double value1, double value2)
static float	geometricMean(float[] values)
static double	geometricMean(java.util.List<? extends java.lang.Number> inputs)
static Pair<java.lang.Double,java.lang.Double>	interQuartileRange(double[] inputs) Calculate the interquartile range.
static Pair<java.lang.Double,java.lang.Double>	interQuartileRange(java.util.List<? extends java.lang.Number> inputs)
static double[]	leverages(double[] inputs) Compute the leverage of each input: (x-xbar)^2/((n-1)sigmax^2))
static float[]	leverages(float[] inputs) Compute the leverage of each input: (x-xbar)^2/((n-1)sigmax^2))
static double	max(double[] values)
static float	max(float[] values)
static double	max(java.util.List<? extends java.lang.Number> values)
static double	mean(double[] inputs) Calculate the mean of the inputs
static float	mean(float[] inputs) Calculate the mean of the inputs
static double	mean(int[] inputs) Calculate the mean of the inputs
static double	mean(java.util.List<? extends java.lang.Number> inputs) Calculate the mean of the inputs
static double	median(double[] inputs) Calculate the median of the inputs.
static float	median(float[] inputs) Calculate the median of the inputs.
static double	median(java.util.List<? extends java.lang.Number> inputs)
static double	min(double[] values)
static double	min(java.util.List<? extends java.lang.Number> values)
static double	percentile(double[] values, double p) Returns an estimate of the pth percentile of the values in the values array, starting with the element in (0-based) position begin in the array and including length values.
static double	percentile(java.util.List<? extends java.lang.Number> values, double p) calculate percentile p of values in list
static double	standardDeviation(double[] inputs) Calculate the standard deviation of the inputs
static float	standardDeviation(float[] inputs) Calculate the standard deviation of the inputs
static double	standardDeviation(java.util.List<? extends java.lang.Number> inputs)
static float	standardDeviationFloatList(java.util.List<? extends java.lang.Number> inputs)
static double	standardGaussian(double x) Calculate a standard Gaussian with mean 0, variance 1
static double[]	studentizedResiduals(double[] xValues, double[] residuals) Convenience method
static double[]	studentizedResiduals(double[] xValues, double[] residuals, double[] leverages) Calculate the studentized residuals of arrays of somethingorother represented here by their X values, leverages (based on X values), and residuals.
static double	sum(double[] values)
static float	sum(float[] values)
static double	sum(java.util.List<? extends java.lang.Number> values)
static double	variance(double[] inputs) Calculate the variance of the inputs
static float	variance(float[] inputs) Calculate the variance of the inputs
static double	weightedGeometricMean(java.util.List<? extends java.lang.Number> inputs, java.util.List<? extends java.lang.Number> weights) Calculate the weighted geometric mean.
static double	weightedMean(java.util.List<? extends java.lang.Number> inputs, java.util.List<? extends java.lang.Number> weights) Calculate the weighted mean

Methods inherited from class java.lang.Object

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

Constructor Detail

BasicStatistics

public BasicStatistics()

Method Detail

standardDeviation

public static float standardDeviation(float[] inputs)

Calculate the standard deviation of the inputs

Parameters:

inputs -

Returns:

standardDeviation

public static double standardDeviation(double[] inputs)

Calculate the standard deviation of the inputs

Parameters:

inputs -

Returns:

standardDeviation

public static double standardDeviation(java.util.List<? extends java.lang.Number> inputs)

standardDeviationFloatList

public static float standardDeviationFloatList(java.util.List<? extends java.lang.Number> inputs)

variance

public static float variance(float[] inputs)

Calculate the variance of the inputs

Parameters:

inputs -

Returns:

variance

public static double variance(double[] inputs)

Calculate the variance of the inputs

Parameters:

inputs -

Returns:

mean

public static double mean(int[] inputs)

Calculate the mean of the inputs

Parameters:

inputs -

Returns:

mean

public static float mean(float[] inputs)

Calculate the mean of the inputs

Parameters:

inputs -

Returns:

mean

public static double mean(java.util.List<? extends java.lang.Number> inputs)

Calculate the mean of the inputs

Parameters:

inputs -

Returns:

mean

public static double mean(double[] inputs)

Calculate the mean of the inputs

Parameters:

inputs -

Returns:

median

public static double median(java.util.List<? extends java.lang.Number> inputs)

median

public static double median(double[] inputs)

Calculate the median of the inputs. This is median as defined in R -- average of the middle two points. NOT necessarily one of the inputs. Clones before sorting, so as not to disrupt input

Parameters:

inputs -

Returns:

interQuartileRange

public static Pair<java.lang.Double,java.lang.Double> interQuartileRange(double[] inputs)

Calculate the interquartile range. I think this is right. Calculate the median, then calculate the median of everything under the median and everything over the median.

Parameters:

inputs -

Returns:

interQuartileRange

public static Pair<java.lang.Double,java.lang.Double> interQuartileRange(java.util.List<? extends java.lang.Number> inputs)

min

public static double min(double[] values)

min

public static double min(java.util.List<? extends java.lang.Number> values)

max

public static double max(double[] values)

max

public static float max(float[] values)

max

public static double max(java.util.List<? extends java.lang.Number> values)

percentile

public static double percentile(java.util.List<? extends java.lang.Number> values,
                                double p)

calculate percentile p of values in list

Parameters:

values -

p -

Returns:

percentile

public static double percentile(double[] values,
                                double p)

Returns an estimate of the pth percentile of the values in the values array, starting with the element in (0-based) position begin in the array and including length values.

Calls to this method do not modify the array Copied from org.apache.commons.math.stat todo: figure out for sure whether having median() call this is OK

Parameters:

values - array of input values

p - the percentile to compute

Returns:

the percentile value

Throws:

java.lang.IllegalArgumentException - if the parameters are not valid or the input array is null

median

public static float median(float[] inputs)

Calculate the median of the inputs. This is median as defined in R -- average of the middle two points. NOT necessarily one of the inputs. Clones before sorting, so as not to disrupt input

Parameters:

inputs -

Returns:

leverages

public static double[] leverages(double[] inputs)

Compute the leverage of each input: (x-xbar)^2/((n-1)sigmax^2))

Parameters:

inputs -

Returns:

leverages

public static float[] leverages(float[] inputs)

Compute the leverage of each input: (x-xbar)^2/((n-1)sigmax^2))

Parameters:

inputs -

Returns:

studentizedResiduals

public static double[] studentizedResiduals(double[] xValues,
                                            double[] residuals)

Convenience method

Parameters:

xValues -

residuals -

Returns:

studentizedResiduals

public static double[] studentizedResiduals(double[] xValues,
                                            double[] residuals,
                                            double[] leverages)

Calculate the studentized residuals of arrays of somethingorother represented here by their X values, leverages (based on X values), and residuals.

Parameters:

xValues -

residuals -

leverages -

Returns:

sum

public static double sum(double[] values)

sum

public static double sum(java.util.List<? extends java.lang.Number> values)

sum

public static float sum(float[] values)

coefficientOfVariation

public static double coefficientOfVariation(double[] input)

coefficientOfVariation

public static double coefficientOfVariation(java.util.List<? extends java.lang.Number> input)

covariance

public static double covariance(double[] xvalues,
                                double[] yvalues)

Covariance = (sum(x*y) - ((sumx)(sumy)/n)) / (n-1)

Parameters:

xvalues -

yvalues -

Returns:

covariance

public static float covariance(float[] xvalues,
                               float[] yvalues)

Covariance = (sum(x*y) - ((sumx)(sumy)/n)) / (n-1)

Parameters:

xvalues -

yvalues -

Returns:

correlationCoefficient

public static double correlationCoefficient(java.util.List<? extends java.lang.Number> xvaluesList,
                                            java.util.List<? extends java.lang.Number> yvaluesList)

correlationCoefficient

public static double correlationCoefficient(double[] xvalues,
                                            double[] yvalues)

covariance(xy) / (sx * sy)

correlationCoefficient

public static float correlationCoefficient(float[] xvalues,
                                           float[] yvalues)

covariance(xy) / (sx * sy)

geometricMean

public static double geometricMean(double[] values)

Geometric mean is defined as ((x1)(x2)(x3)...(xk)) ^ 1/k

Parameters:

values -

Returns:

geometricMean

public static double geometricMean(java.util.List<? extends java.lang.Number> inputs)

geometricMean

public static float geometricMean(float[] values)

geometricMean

public static double geometricMean(double value1,
                                   double value2)

weightedGeometricMean

public static double weightedGeometricMean(java.util.List<? extends java.lang.Number> inputs,
                                           java.util.List<? extends java.lang.Number> weights)

Calculate the weighted geometric mean. Exponentiated weighted mean of logs.

Parameters:

inputs -

weights -

Returns:

weightedMean

public static double weightedMean(java.util.List<? extends java.lang.Number> inputs,
                                  java.util.List<? extends java.lang.Number> weights)

Calculate the weighted mean

Parameters:

inputs -

weights -

Returns:

standardGaussian

public static double standardGaussian(double x)

Calculate a standard Gaussian with mean 0, variance 1

Parameters:

x -

Returns:

calcNormalDistDensity

public static double calcNormalDistDensity(double mu,
                                           double sigma,
                                           double x)

Calculate the density of the normal distribution function with parameters mu and sigma, at point x

Parameters:

mu -

sigma -

x -

Returns:

erf

public static double erf(double z)

Error function for standard normal distribution

Parameters:

z -

Returns:

calcStandardNormalCumDensity

public static double calcStandardNormalCumDensity(double z)

Calculate the cumulative density of the standard normal distribution at point x

Parameters:

z -

Returns:

calcNormalCumDensity

public static double calcNormalCumDensity(double mu,
                                          double sigma,
                                          double x)

Calculate the cumulative density of the normal distribution function with parameters mu and sigma, at point x

Parameters:

mu -

sigma -

x -

Returns: