221 lines
8.2 KiB
C++
221 lines
8.2 KiB
C++
/* -*- Mode: C++; tab-width: 4; indent-tabs-mode: nil; c-basic-offset: 4 -*- */
|
|
/*
|
|
* This file is part of the LibreOffice project.
|
|
*
|
|
* This Source Code Form is subject to the terms of the Mozilla Public
|
|
* License, v. 2.0. If a copy of the MPL was not distributed with this
|
|
* file, You can obtain one at http://mozilla.org/MPL/2.0/.
|
|
*
|
|
* This file incorporates work covered by the following license notice:
|
|
*
|
|
* Licensed to the Apache Software Foundation (ASF) under one or more
|
|
* contributor license agreements. See the NOTICE file distributed
|
|
* with this work for additional information regarding copyright
|
|
* ownership. The ASF licenses this file to you under the Apache
|
|
* License, Version 2.0 (the "License"); you may not use this file
|
|
* except in compliance with the License. You may obtain a copy of
|
|
* the License at http://www.apache.org/licenses/LICENSE-2.0 .
|
|
*/
|
|
|
|
#include <sal/config.h>
|
|
|
|
#include <limits>
|
|
#include <string_view>
|
|
|
|
#include <ExponentialRegressionCurveCalculator.hxx>
|
|
#include <RegressionCalculationHelper.hxx>
|
|
#include <SpecialCharacters.hxx>
|
|
|
|
#include <rtl/math.hxx>
|
|
#include <rtl/ustrbuf.hxx>
|
|
|
|
using namespace ::com::sun::star;
|
|
|
|
namespace chart
|
|
{
|
|
|
|
ExponentialRegressionCurveCalculator::ExponentialRegressionCurveCalculator()
|
|
: m_fLogSlope(std::numeric_limits<double>::quiet_NaN())
|
|
, m_fLogIntercept(std::numeric_limits<double>::quiet_NaN())
|
|
, m_fSign(1.0)
|
|
{
|
|
}
|
|
|
|
ExponentialRegressionCurveCalculator::~ExponentialRegressionCurveCalculator()
|
|
{}
|
|
|
|
// ____ XRegressionCurveCalculator ____
|
|
void SAL_CALL ExponentialRegressionCurveCalculator::recalculateRegression(
|
|
const uno::Sequence< double >& aXValues,
|
|
const uno::Sequence< double >& aYValues )
|
|
{
|
|
RegressionCalculationHelper::tDoubleVectorPair aValues(
|
|
RegressionCalculationHelper::cleanup(
|
|
aXValues, aYValues,
|
|
RegressionCalculationHelper::isValidAndYPositive()));
|
|
m_fSign = 1.0;
|
|
|
|
size_t nMax = aValues.first.size();
|
|
if( nMax <= 1 ) // at least 2 points
|
|
{
|
|
aValues = RegressionCalculationHelper::cleanup(
|
|
aXValues, aYValues,
|
|
RegressionCalculationHelper::isValidAndYNegative());
|
|
nMax = aValues.first.size();
|
|
if( nMax <= 1 )
|
|
{
|
|
m_fLogSlope = std::numeric_limits<double>::quiet_NaN();
|
|
m_fLogIntercept = std::numeric_limits<double>::quiet_NaN();
|
|
m_fCorrelationCoefficient = std::numeric_limits<double>::quiet_NaN();// actual it is coefficient of determination
|
|
return;
|
|
}
|
|
m_fSign = -1.0;
|
|
}
|
|
|
|
double fAverageX = 0.0, fAverageY = 0.0;
|
|
double fLogIntercept = ( mForceIntercept && (m_fSign * mInterceptValue)>0 ) ? log(m_fSign * mInterceptValue) : 0.0;
|
|
std::vector<double> yVector;
|
|
yVector.resize(nMax, 0.0);
|
|
|
|
size_t i = 0;
|
|
for( i = 0; i < nMax; ++i )
|
|
{
|
|
double yValue = log( m_fSign *aValues.second[i] );
|
|
if (mForceIntercept)
|
|
{
|
|
yValue -= fLogIntercept;
|
|
}
|
|
else
|
|
{
|
|
fAverageX += aValues.first[i];
|
|
fAverageY += yValue;
|
|
}
|
|
yVector[i] = yValue;
|
|
}
|
|
|
|
const double fN = static_cast< double >( nMax );
|
|
fAverageX /= fN;
|
|
fAverageY /= fN;
|
|
|
|
double fQx = 0.0, fQy = 0.0, fQxy = 0.0;
|
|
for( i = 0; i < nMax; ++i )
|
|
{
|
|
double fDeltaX = aValues.first[i] - fAverageX;
|
|
double fDeltaY = yVector[i] - fAverageY;
|
|
|
|
fQx += fDeltaX * fDeltaX;
|
|
fQy += fDeltaY * fDeltaY;
|
|
fQxy += fDeltaX * fDeltaY;
|
|
}
|
|
|
|
m_fLogSlope = fQxy / fQx;
|
|
m_fLogIntercept = mForceIntercept ? fLogIntercept : fAverageY - m_fLogSlope * fAverageX;
|
|
m_fCorrelationCoefficient = fQxy / sqrt( fQx * fQy );
|
|
}
|
|
|
|
double SAL_CALL ExponentialRegressionCurveCalculator::getCurveValue( double x )
|
|
{
|
|
if( ! ( std::isnan( m_fLogSlope ) ||
|
|
std::isnan( m_fLogIntercept )))
|
|
{
|
|
return m_fSign * exp(m_fLogIntercept + x * m_fLogSlope);
|
|
}
|
|
|
|
return std::numeric_limits<double>::quiet_NaN();
|
|
}
|
|
|
|
uno::Sequence< geometry::RealPoint2D > SAL_CALL ExponentialRegressionCurveCalculator::getCurveValues(
|
|
double min, double max, ::sal_Int32 nPointCount,
|
|
const uno::Reference< chart2::XScaling >& xScalingX,
|
|
const uno::Reference< chart2::XScaling >& xScalingY,
|
|
sal_Bool bMaySkipPointsInCalculation )
|
|
{
|
|
if( bMaySkipPointsInCalculation &&
|
|
isLinearScaling( xScalingX ) &&
|
|
isLogarithmicScaling( xScalingY ))
|
|
{
|
|
// optimize result
|
|
uno::Sequence< geometry::RealPoint2D > aResult{ { min, getCurveValue( min ) },
|
|
{ max, getCurveValue( max ) } };
|
|
|
|
return aResult;
|
|
}
|
|
|
|
return RegressionCurveCalculator::getCurveValues( min, max, nPointCount, xScalingX, xScalingY, bMaySkipPointsInCalculation );
|
|
}
|
|
|
|
OUString ExponentialRegressionCurveCalculator::ImplGetRepresentation(
|
|
const uno::Reference< util::XNumberFormatter >& xNumFormatter,
|
|
sal_Int32 nNumberFormatKey, sal_Int32* pFormulaMaxWidth /* = nullptr */ ) const
|
|
{
|
|
double fIntercept = exp(m_fLogIntercept);
|
|
bool bHasSlope = !rtl::math::approxEqual( exp(m_fLogSlope), 1.0 );
|
|
bool bHasLogSlope = !rtl::math::approxEqual( fabs(m_fLogSlope), 1.0 );
|
|
bool bHasIntercept = !rtl::math::approxEqual( fIntercept, 1.0 ) && fIntercept != 0.0;
|
|
|
|
OUStringBuffer aBuf( mYName + " = " );
|
|
sal_Int32 nLineLength = aBuf.getLength();
|
|
sal_Int32 nValueLength=0;
|
|
if ( pFormulaMaxWidth && *pFormulaMaxWidth > 0 )
|
|
{ // count characters different from coefficients
|
|
sal_Int32 nCharMin = nLineLength + 10 + mXName.getLength(); // 10 = "exp( ", " x )" + 2 extra characters
|
|
if ( m_fSign < 0.0 )
|
|
nCharMin += 2;
|
|
if ( fIntercept == 0.0 || ( !bHasSlope && m_fLogIntercept != 0.0 ) )
|
|
nCharMin += 3; // " + " special case where equation is written exp( a + b x )
|
|
if ( ( bHasIntercept || fIntercept == 0.0 || ( !bHasSlope && m_fLogIntercept != 0.0 ) ) &&
|
|
bHasLogSlope )
|
|
nValueLength = ( *pFormulaMaxWidth - nCharMin ) / 2;
|
|
else
|
|
nValueLength = *pFormulaMaxWidth - nCharMin;
|
|
if ( nValueLength <= 0 )
|
|
nValueLength = 1;
|
|
}
|
|
// temporary buffer
|
|
OUStringBuffer aTmpBuf("");
|
|
// if nValueLength not calculated then nullptr
|
|
sal_Int32* pValueLength = nValueLength ? &nValueLength : nullptr;
|
|
if ( m_fSign < 0.0 )
|
|
aTmpBuf.append( OUStringChar(aMinusSign) + " " );
|
|
if ( bHasIntercept )
|
|
{
|
|
OUString aValueString = getFormattedString( xNumFormatter, nNumberFormatKey, fIntercept, pValueLength );
|
|
if ( aValueString != "1" ) // aValueString may be rounded to 1 if nValueLength is small
|
|
{
|
|
aTmpBuf.append( aValueString + " " );
|
|
addStringToEquation( aBuf, nLineLength, aTmpBuf, pFormulaMaxWidth );
|
|
aTmpBuf.truncate();
|
|
}
|
|
}
|
|
aTmpBuf.append( "exp( " );
|
|
if ( !bHasIntercept )
|
|
{
|
|
if ( fIntercept == 0.0 || // underflow, a true zero is impossible
|
|
( !bHasSlope && m_fLogIntercept != 0.0 ) ) // show logarithmic output, if intercept and slope both are near one
|
|
{ // otherwise drop output of intercept, which is 1 here
|
|
OUString aValueString = getFormattedString( xNumFormatter, nNumberFormatKey, m_fLogIntercept, pValueLength );
|
|
if ( aValueString != "0" ) // aValueString may be rounded to 0 if nValueLength is small
|
|
{
|
|
aTmpBuf.append( aValueString ).append( (m_fLogSlope < 0.0) ? std::u16string_view(u" ") : std::u16string_view(u" + ") );
|
|
}
|
|
}
|
|
}
|
|
if ( m_fLogSlope < 0.0 )
|
|
aTmpBuf.append( OUStringChar(aMinusSign) + " " );
|
|
if ( bHasLogSlope )
|
|
{
|
|
OUString aValueString = getFormattedString( xNumFormatter, nNumberFormatKey, fabs(m_fLogSlope), pValueLength );
|
|
if ( aValueString != "1" ) // aValueString may be rounded to 1 if nValueLength is small
|
|
{
|
|
aTmpBuf.append( aValueString + " " );
|
|
}
|
|
}
|
|
aTmpBuf.append( mXName + " )");
|
|
addStringToEquation( aBuf, nLineLength, aTmpBuf, pFormulaMaxWidth );
|
|
|
|
return aBuf.makeStringAndClear();
|
|
}
|
|
|
|
} // namespace chart
|
|
|
|
/* vim:set shiftwidth=4 softtabstop=4 expandtab: */
|