/*
// $Id: //open/mondrian-release/3.2/src/main/mondrian/olap/fun/CovarianceFunDef.java#1 $
// This software is subject to the terms of the Eclipse Public License v1.0
// Agreement, available at the following URL:
// http://www.eclipse.org/legal/epl-v10.html.
// Copyright (C) 2006-2009 Julian Hyde
// All Rights Reserved.
// You must accept the terms of that agreement to use this software.
*/
package mondrian.olap.fun;
import mondrian.olap.*;
import mondrian.calc.Calc;
import mondrian.calc.ExpCompiler;
import mondrian.calc.ListCalc;
import mondrian.calc.impl.ValueCalc;
import mondrian.calc.impl.AbstractDoubleCalc;
import mondrian.mdx.ResolvedFunCall;
import java.util.List;
/**
* Definition of the <code>Covariance</code> and
* <code>CovarianceN</code> MDX functions.
*
* @author jhyde
* @version $Id: //open/mondrian-release/3.2/src/main/mondrian/olap/fun/CovarianceFunDef.java#1 $
* @since Mar 23, 2006
*/
class CovarianceFunDef extends FunDefBase {
static final ReflectiveMultiResolver CovarianceResolver =
new ReflectiveMultiResolver(
"Covariance",
"Covariance(<Set>, <Numeric Expression>[, <Numeric Expression>])",
"Returns the covariance of two series evaluated over a set (biased).",
new String[]{"fnxn", "fnxnn"},
CovarianceFunDef.class);
static final MultiResolver CovarianceNResolver =
new ReflectiveMultiResolver(
"CovarianceN",
"CovarianceN(<Set>, <Numeric Expression>[, <Numeric Expression>])",
"Returns the covariance of two series evaluated over a set (unbiased).",
new String[]{"fnxn", "fnxnn"},
CovarianceFunDef.class);
private final boolean biased;
public CovarianceFunDef(FunDef dummyFunDef) {
super(dummyFunDef);
this.biased = dummyFunDef.getName().equals("Covariance");
}
public Calc compileCall(ResolvedFunCall call, ExpCompiler compiler) {
final ListCalc listCalc =
compiler.compileList(call.getArg(0));
final Calc calc1 =
compiler.compileScalar(call.getArg(1), true);
final Calc calc2 =
call.getArgCount() > 2
? compiler.compileScalar(call.getArg(2), true)
: new ValueCalc(call);
return new AbstractDoubleCalc(call, new Calc[] {listCalc, calc1, calc2})
{
public double evaluateDouble(Evaluator evaluator) {
List memberList = listCalc.evaluateList(evaluator);
return (Double) covariance(
evaluator.push(false), memberList, calc1, calc2, biased);
}
public boolean dependsOn(Hierarchy hierarchy) {
return anyDependsButFirst(getCalcs(), hierarchy);
}
};
}
}
// End CovarianceFunDef.java