Line data Source code
1 : /* -*- Mode: C++; tab-width: 4; indent-tabs-mode: nil; c-basic-offset: 4 -*- */
2 : /*
3 : * This file is part of the LibreOffice project.
4 : *
5 : * This Source Code Form is subject to the terms of the Mozilla Public
6 : * License, v. 2.0. If a copy of the MPL was not distributed with this
7 : * file, You can obtain one at http://mozilla.org/MPL/2.0/.
8 : *
9 : * This file incorporates work covered by the following license notice:
10 : *
11 : * Licensed to the Apache Software Foundation (ASF) under one or more
12 : * contributor license agreements. See the NOTICE file distributed
13 : * with this work for additional information regarding copyright
14 : * ownership. The ASF licenses this file to you under the Apache
15 : * License, Version 2.0 (the "License"); you may not use this file
16 : * except in compliance with the License. You may obtain a copy of
17 : * the License at http://www.apache.org/licenses/LICENSE-2.0 .
18 : */
19 :
20 : #include "LogarithmicRegressionCurveCalculator.hxx"
21 : #include "macros.hxx"
22 : #include "RegressionCalculationHelper.hxx"
23 :
24 : #include <rtl/math.hxx>
25 : #include <rtl/ustrbuf.hxx>
26 :
27 : using namespace ::com::sun::star;
28 :
29 : using ::rtl::OUString;
30 : using ::rtl::OUStringBuffer;
31 :
32 : namespace chart
33 : {
34 :
35 0 : LogarithmicRegressionCurveCalculator::LogarithmicRegressionCurveCalculator() :
36 : m_fSlope( 0.0 ),
37 0 : m_fIntercept( 0.0 )
38 : {
39 0 : ::rtl::math::setNan( & m_fSlope );
40 0 : ::rtl::math::setNan( & m_fIntercept );
41 0 : }
42 :
43 0 : LogarithmicRegressionCurveCalculator::~LogarithmicRegressionCurveCalculator()
44 0 : {}
45 :
46 : // ____ XRegressionCurve ____
47 0 : void SAL_CALL LogarithmicRegressionCurveCalculator::recalculateRegression(
48 : const uno::Sequence< double >& aXValues,
49 : const uno::Sequence< double >& aYValues )
50 : throw (uno::RuntimeException)
51 : {
52 : RegressionCalculationHelper::tDoubleVectorPair aValues(
53 : RegressionCalculationHelper::cleanup(
54 : aXValues, aYValues,
55 0 : RegressionCalculationHelper::isValidAndXPositive()));
56 :
57 0 : const size_t nMax = aValues.first.size();
58 0 : if( nMax == 0 )
59 : {
60 0 : ::rtl::math::setNan( & m_fSlope );
61 0 : ::rtl::math::setNan( & m_fIntercept );
62 0 : ::rtl::math::setNan( & m_fCorrelationCoeffitient );
63 0 : return;
64 : }
65 :
66 0 : double fAverageX = 0.0, fAverageY = 0.0;
67 0 : size_t i = 0;
68 0 : for( i = 0; i < nMax; ++i )
69 : {
70 0 : fAverageX += log( aValues.first[i] );
71 0 : fAverageY += aValues.second[i];
72 : }
73 :
74 0 : const double fN = static_cast< double >( nMax );
75 0 : fAverageX /= fN;
76 0 : fAverageY /= fN;
77 :
78 0 : double fQx = 0.0, fQy = 0.0, fQxy = 0.0;
79 0 : for( i = 0; i < nMax; ++i )
80 : {
81 0 : double fDeltaX = log( aValues.first[i] ) - fAverageX;
82 0 : double fDeltaY = aValues.second[i] - fAverageY;
83 :
84 0 : fQx += fDeltaX * fDeltaX;
85 0 : fQy += fDeltaY * fDeltaY;
86 0 : fQxy += fDeltaX * fDeltaY;
87 : }
88 :
89 0 : m_fSlope = fQxy / fQx;
90 0 : m_fIntercept = fAverageY - m_fSlope * fAverageX;
91 0 : m_fCorrelationCoeffitient = fQxy / sqrt( fQx * fQy );
92 : }
93 :
94 0 : double SAL_CALL LogarithmicRegressionCurveCalculator::getCurveValue( double x )
95 : throw (lang::IllegalArgumentException,
96 : uno::RuntimeException)
97 : {
98 : double fResult;
99 0 : ::rtl::math::setNan( & fResult );
100 :
101 0 : if( ! ( ::rtl::math::isNan( m_fSlope ) ||
102 0 : ::rtl::math::isNan( m_fIntercept )))
103 : {
104 0 : fResult = m_fSlope * log( x ) + m_fIntercept;
105 : }
106 :
107 0 : return fResult;
108 : }
109 :
110 0 : uno::Sequence< geometry::RealPoint2D > SAL_CALL LogarithmicRegressionCurveCalculator::getCurveValues(
111 : double min, double max, ::sal_Int32 nPointCount,
112 : const uno::Reference< chart2::XScaling >& xScalingX,
113 : const uno::Reference< chart2::XScaling >& xScalingY,
114 : ::sal_Bool bMaySkipPointsInCalculation )
115 : throw (lang::IllegalArgumentException,
116 : uno::RuntimeException)
117 : {
118 0 : if( bMaySkipPointsInCalculation &&
119 0 : isLogarithmicScaling( xScalingX ) &&
120 0 : isLinearScaling( xScalingY ))
121 : {
122 : // optimize result
123 0 : uno::Sequence< geometry::RealPoint2D > aResult( 2 );
124 0 : aResult[0].X = min;
125 0 : aResult[0].Y = this->getCurveValue( min );
126 0 : aResult[1].X = max;
127 0 : aResult[1].Y = this->getCurveValue( max );
128 :
129 0 : return aResult;
130 : }
131 0 : return RegressionCurveCalculator::getCurveValues( min, max, nPointCount, xScalingX, xScalingY, bMaySkipPointsInCalculation );
132 : }
133 :
134 0 : OUString LogarithmicRegressionCurveCalculator::ImplGetRepresentation(
135 : const uno::Reference< util::XNumberFormatter >& xNumFormatter,
136 : ::sal_Int32 nNumberFormatKey ) const
137 : {
138 0 : OUStringBuffer aBuf( C2U( "f(x) = " ));
139 :
140 0 : bool bHaveSlope = false;
141 :
142 0 : if( m_fSlope != 0.0 )
143 : {
144 0 : if( ::rtl::math::approxEqual( fabs( m_fSlope ), 1.0 ))
145 : {
146 0 : if( m_fSlope < 0 )
147 0 : aBuf.append( UC_MINUS_SIGN );
148 : }
149 : else
150 : {
151 0 : aBuf.append( getFormattedString( xNumFormatter, nNumberFormatKey, m_fSlope ));
152 0 : aBuf.append( UC_SPACE );
153 : }
154 0 : aBuf.appendAscii( RTL_CONSTASCII_STRINGPARAM( "ln(x)" ));
155 0 : bHaveSlope = true;
156 : }
157 :
158 0 : if( bHaveSlope )
159 : {
160 0 : if( m_fIntercept < 0.0 )
161 : {
162 0 : aBuf.append( UC_SPACE );
163 0 : aBuf.append( UC_MINUS_SIGN );
164 0 : aBuf.append( UC_SPACE );
165 0 : aBuf.append( getFormattedString( xNumFormatter, nNumberFormatKey, fabs( m_fIntercept )));
166 : }
167 0 : else if( m_fIntercept > 0.0 )
168 : {
169 0 : aBuf.appendAscii( RTL_CONSTASCII_STRINGPARAM( " + " ));
170 0 : aBuf.append( getFormattedString( xNumFormatter, nNumberFormatKey, m_fIntercept ));
171 : }
172 : }
173 : else
174 : {
175 0 : aBuf.append( getFormattedString( xNumFormatter, nNumberFormatKey, m_fIntercept ));
176 : }
177 :
178 0 : return aBuf.makeStringAndClear();
179 : }
180 :
181 : } // namespace chart
182 :
183 : /* vim:set shiftwidth=4 softtabstop=4 expandtab: */
|
