Package org.apache.poi.ss.formula.functions

Class Irr

java.lang.Object

org.apache.poi.ss.formula.functions.Irr

All Implemented Interfaces:

Function

public final class Irr
extends Object
implements Function

Calculates the internal rate of return. Syntax is IRR(values) or IRR(values,guess)

See Also:

Wikipedia on IRR, Excel IRR

Constructor Summary

Constructors

Constructor	Description
Irr()

Method Summary

Modifier and Type	Method	Description
ValueEval	evaluate(ValueEval[] args, int srcRowIndex, int srcColumnIndex)
static double	irr(double[] income)	Computes the internal rate of return using an estimated irr of 10 percent.
static double	irr(double[] values, double guess)	Calculates IRR using the Newton-Raphson Method.

Methods inherited from class java.lang.Object

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

Constructor Detail

Irr

public Irr()

Method Detail

evaluate

public ValueEval evaluate(ValueEval[] args,
                          int srcRowIndex,
                          int srcColumnIndex)

Specified by:

evaluate in interface Function

Parameters:

args - the evaluated function arguments. Empty values are represented with BlankEval or MissingArgEval, never null.

srcRowIndex - row index of the cell containing the formula under evaluation

srcColumnIndex - column index of the cell containing the formula under evaluation

Returns:

The evaluated result, possibly an ErrorEval, never null. Note - Excel uses the error code #NUM! instead of IEEE NaN, so when numeric functions evaluate to Double.NaN be sure to translate the result to ErrorEval.NUM_ERROR.

irr

public static double irr(double[] income)

Computes the internal rate of return using an estimated irr of 10 percent.

Parameters:

income - the income values.

Returns:

the irr.

irr

public static double irr(double[] values,
                         double guess)

Calculates IRR using the Newton-Raphson Method.

Starting with the guess, the method cycles through the calculation until the result is accurate within 0.00001 percent. If IRR can't find a result that works after 1000 tries, the Double.NaN is returned.

The implementation is inspired by the NewtonSolver from the Apache Commons-Math library,

Parameters:

values - the income values.

guess - the initial guess of irr.

Returns:

the irr value. The method returns Double.NaN if the maximum iteration count is exceeded

See Also:

http://commons.apache.org, http://en.wikipedia.org/wiki/Internal_rate_of_return#Numerical_solution, http://en.wikipedia.org/wiki/Newton%27s_method