1<?xml version="1.0" encoding="UTF-8"?> 2<helpdocument version="1.0"> 3 4<!-- 5*********************************************************************** 6 * 7 * DO NOT ALTER OR REMOVE COPYRIGHT NOTICES OR THIS FILE HEADER. 8 * 9 * Copyright 2000, 2010 Oracle and/or its affiliates. 10 * 11 * OpenOffice.org - a multi-platform office productivity suite 12 * 13 * This file is part of OpenOffice.org. 14 * 15 * OpenOffice.org is free software: you can redistribute it and/or modify 16 * it under the terms of the GNU Lesser General Public License version 3 17 * only, as published by the Free Software Foundation. 18 * 19 * OpenOffice.org is distributed in the hope that it will be useful, 20 * but WITHOUT ANY WARRANTY; without even the implied warranty of 21 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the 22 * GNU Lesser General Public License version 3 for more details 23 * (a copy is included in the LICENSE file that accompanied this code). 24 * 25 * You should have received a copy of the GNU Lesser General Public License 26 * version 3 along with OpenOffice.org. If not, see 27 * <http://www.openoffice.org/license.html> 28 * for a copy of the LGPLv3 License. 29 * 30 ************************************************************************ 31 --> 32 33 34<meta> 35 <topic id="textschart0104050100xml" indexer="include" status="PUBLISH"> 36 <title xml-lang="en-US" id="tit">Trend Lines </title> 37 <filename>/text/schart/01/04050100.xhp</filename> 38 </topic> 39 </meta> 40 <body> 41<bookmark xml-lang="en-US" branch="index" id="bm_id1744743"><bookmark_value>calculating;regression curves</bookmark_value> 42 <bookmark_value>regression curves in charts</bookmark_value> 43 <bookmark_value>trend lines in charts</bookmark_value> 44 <bookmark_value>mean value lines in charts</bookmark_value> 45</bookmark> 46<comment>only use trend line, not regression curve: i89823</comment> 47<bookmark xml-lang="en-US" branch="hid/.uno:InsertMenuTrendlines" id="bm_id2001709" localize="false"/> 48<paragraph xml-lang="en-US" id="hd_id5409405" role="heading" level="1" l10n="CHG"><variable id="regression"><link href="text/schart/01/04050100.xhp">Trend Lines</link> 49</variable></paragraph> 50 <paragraph xml-lang="en-US" id="par_id7272255" role="paragraph" l10n="CHG"><variable id="trendlinestext"><ahelp hid=".">Trend lines can be added to all 2D chart types except for Pie and Stock charts.</ahelp> 51</variable></paragraph> 52 <embed href="text/schart/01/04030000.xhp#context"/> 53 <section id="howtoget"> 54 <embed href="text/schart/00/00000004.xhp#trendlines"/> 55 </section><comment>None</comment> 56<bookmark xml-lang="en-US" branch="hid/CHART2_HID_SCH_TRENDLINE_RB_NONE" id="bm_id2659077" localize="false"/> 57<paragraph xml-lang="en-US" id="par_id143436" role="paragraph" l10n="NEW"><ahelp hid="." visibility="hidden">No trend line is shown.</ahelp></paragraph><comment>Linear</comment> 58<bookmark xml-lang="en-US" branch="hid/CHART2_HID_SCH_TRENDLINE_RB_LINEAR" id="bm_id6944181" localize="false"/> 59<paragraph xml-lang="en-US" id="par_id5716727" role="paragraph" l10n="NEW"><ahelp hid="." visibility="hidden">A linear trend line is shown.</ahelp></paragraph><comment>Logarithmic</comment> 60<bookmark xml-lang="en-US" branch="hid/CHART2_HID_SCH_TRENDLINE_RB_LOGARITHMIC" id="bm_id886257" localize="false"/> 61<paragraph xml-lang="en-US" id="par_id5840021" role="paragraph" l10n="NEW"><ahelp hid="." visibility="hidden">A logarithmic trend line is shown.</ahelp></paragraph><comment>Exponential</comment> 62<bookmark xml-lang="en-US" branch="hid/CHART2_HID_SCH_TRENDLINE_RB_EXPONENTIAL" id="bm_id8892483" localize="false"/> 63<paragraph xml-lang="en-US" id="par_id9417096" role="paragraph" l10n="NEW"><ahelp hid="." visibility="hidden">An exponential trend line is shown.</ahelp></paragraph><comment>Power</comment> 64<bookmark xml-lang="en-US" branch="hid/CHART2_HID_SCH_TRENDLINE_RB_POWER" id="bm_id7198400" localize="false"/> 65<paragraph xml-lang="en-US" id="par_id8482924" role="paragraph" l10n="NEW"><ahelp hid="." visibility="hidden">A power trend line is shown.</ahelp></paragraph><comment>Show equation</comment> 66<bookmark xml-lang="en-US" branch="hid/CHART2_HID_SCH_TRENDLINE_SHOW_EQUATION" id="bm_id4724875" localize="false"/> 67<paragraph xml-lang="en-US" id="par_id8962370" role="paragraph" l10n="NEW"><ahelp hid="." visibility="hidden">Shows the trend line equation next to the trend line.</ahelp></paragraph><comment>Show correlation coefficient (R2)</comment> 68<bookmark xml-lang="en-US" branch="hid/CHART2_HID_SCH_TRENDLINE_SHOW_R_SQUARED" id="bm_id9068880" localize="false"/> 69<paragraph xml-lang="en-US" id="par_id6889858" role="paragraph" l10n="CHG"><ahelp hid="." visibility="hidden">Shows the coefficient of determination next to the trend line.</ahelp></paragraph> 70 <paragraph xml-lang="en-US" id="par_id8398998" role="note" l10n="CHG">If you insert a trend line to a chart type that uses categories, like <emph>Line </emph>or <emph>Column, </emph>then the numbers 1, 2, 3, <emph>…</emph> are used as x-values to calculate the trend line.</paragraph> 71 <list type="ordered"> 72 <listitem> 73 <paragraph xml-lang="en-US" id="par_id5676747" role="paragraph" l10n="CHG">To insert trend lines for all data series, double-click the chart to enter edit mode. Choose <item type="menuitem">Insert - Trend Lines</item>, then select the type of trend line from None, Linear, Logarithmic, Exponential, or Power trend line.</paragraph> 74 </listitem> 75 <listitem> 76 <paragraph xml-lang="en-US" id="par_id4349192" role="paragraph" l10n="CHG">To insert a trend line for a single data series, select the data series in the chart, right-click to open the context menu, and choose <item type="menuitem">Insert - Trend Line</item>.</paragraph> 77 </listitem> 78 <listitem> 79 <paragraph xml-lang="en-US" id="par_id9337443" role="paragraph" l10n="NEW">To delete a single trend line or mean value line, click the line, then press the Del key.</paragraph> 80 </listitem> 81 <listitem> 82 <paragraph xml-lang="en-US" id="par_id4529251" role="paragraph" l10n="NEW">To delete all trend lines, choose <item type="menuitem">Insert - Trend Lines</item>, then select <emph>None</emph>.</paragraph> 83 </listitem> 84 </list> 85 <paragraph xml-lang="en-US" id="par_id296334" role="note" l10n="CHG">A trend line is shown in the legend automatically.</paragraph> 86<bookmark xml-lang="en-US" branch="hid/.uno:InsertMenuMeanValues" id="bm_id7854243" localize="false"/> 87<paragraph xml-lang="en-US" id="par_id4072084" role="paragraph" l10n="NEW"><ahelp hid=".">Mean Value Lines are special trend lines that show the mean value. Use <item type="menuitem">Insert - Mean Value Lines</item> to insert mean value lines for data series.</ahelp></paragraph> 88<embed href="text/schart/01/04030000.xhp#context"/> 89 <paragraph xml-lang="en-US" id="par_id9569689" role="paragraph" l10n="NEW">The trend line has the same color as the corresponding data series. To change the line properties, select the trend line and choose <item type="menuitem">Format - Format Selection - Line</item>.</paragraph> 90<bookmark xml-lang="en-US" branch="hid/.uno:InsertTrendlineEquation" id="bm_id2754600" localize="false"/> 91<paragraph xml-lang="en-US" id="par_id846888" role="paragraph" l10n="NEW"><ahelp hid=".">To show the trend line equation, select the trend line in the chart, right-click to open the context menu, and choose <emph>Insert Trend Line Equation</emph>.</ahelp></paragraph> 92 <paragraph xml-lang="en-US" id="par_id8962065" role="paragraph" l10n="CHG">When the chart is in edit mode, %PRODUCTNAME gives you the equation of the trend line and the coefficient of determination R². Click on the trend line to see the information in the status bar.</paragraph> 93 <paragraph xml-lang="en-US" id="par_id1328470" role="note" l10n="CHG">For a category chart (for example a line chart), the trend line information is calculated using numbers 1, 2, 3, … as x-values. This is also true if your data series uses other numbers as names for the x-values. For such charts the XY chart type might be more suitable.</paragraph> 94 <paragraph xml-lang="en-US" id="par_id8092593" role="paragraph" l10n="CHG">To show the equation and the coefficient of determination, select the trend line and choose <item type="menuitem">Format - Format Selection - Equation</item>.</paragraph> 95<paragraph xml-lang="en-US" id="par_id7971434" role="paragraph" l10n="CHG"><ahelp hid="." visibility="hidden">Enable Show equation to see the equation of the trend line.</ahelp></paragraph><comment>hid</comment> 96<paragraph xml-lang="en-US" id="par_id558793" role="paragraph" l10n="CHG"><ahelp hid="." visibility="hidden">Enable Show Coefficient of Determination to see the determination coefficient of the trend line.</ahelp></paragraph> 97 <paragraph xml-lang="en-US" id="par_id7735221" role="paragraph" l10n="NEW">You can also calculate the parameters using Calc functions as follows.</paragraph> 98 <paragraph xml-lang="en-US" id="hd_id5744193" role="heading" level="2" l10n="NEW">The linear regression equation</paragraph> 99 <paragraph xml-lang="en-US" id="par_id9251991" role="paragraph" l10n="NEW">The <emph>linear regression</emph> follows the equation <item type="literal">y=m*x+b</item>.</paragraph> 100 <paragraph xml-lang="en-US" id="par_id7951902" role="code" l10n="NEW">m = SLOPE(Data_Y;Data_X) </paragraph> 101 <paragraph xml-lang="en-US" id="par_id6637165" role="code" l10n="NEW">b = INTERCEPT(Data_Y ;Data_X) </paragraph> 102 <paragraph xml-lang="en-US" id="par_id7879268" role="paragraph" l10n="NEW">Calculate the coefficient of determination by</paragraph> 103 <paragraph xml-lang="en-US" id="par_id9244361" role="code" l10n="NEW">r² = RSQ(Data_Y;Data_X) </paragraph> 104 <paragraph xml-lang="en-US" id="par_id2083498" role="paragraph" l10n="NEW">Besides m, b and r² the array function <emph>LINEST</emph> provides additional statistics for a regression analysis.</paragraph> 105 <paragraph xml-lang="en-US" id="hd_id2538834" role="heading" level="2" l10n="NEW">The logarithm regression equation</paragraph> 106 <paragraph xml-lang="en-US" id="par_id394299" role="paragraph" l10n="NEW">The <emph>logarithm regression</emph> follows the equation <item type="literal">y=a*ln(x)+b</item>.</paragraph> 107 <paragraph xml-lang="en-US" id="par_id2134159" role="code" l10n="NEW">a = SLOPE(Data_Y;LN(Data_X)) </paragraph> 108 <paragraph xml-lang="en-US" id="par_id5946531" role="code" l10n="NEW">b = INTERCEPT(Data_Y ;LN(Data_X)) </paragraph> 109 <paragraph xml-lang="en-US" id="par_id5649281" role="code" l10n="NEW">r² = RSQ(Data_Y;LN(Data_X)) </paragraph> 110 <paragraph xml-lang="en-US" id="hd_id7874080" role="heading" level="2" l10n="NEW">The exponential regression equation</paragraph> 111 <paragraph xml-lang="en-US" id="par_id4679097" role="paragraph" l10n="CHG"> For exponential trend lines a transformation to a linear model takes place. The optimal curve fitting is related to the linear model and the results are interpreted accordingly. </paragraph> 112 <paragraph xml-lang="en-US" id="par_id9112216" role="paragraph" l10n="NEW">The exponential regression follows the equation <item type="literal">y=b*exp(a*x)</item> or <item type="literal">y=b*m^x</item>, which is transformed to <item type="literal">ln(y)=ln(b)+a*x</item> or <item type="literal">ln(y)=ln(b)+ln(m)*x</item> respectively.</paragraph> 113 <paragraph xml-lang="en-US" id="par_id4416638" role="code" l10n="NEW">a = SLOPE(LN(Data_Y);Data_X) </paragraph> 114 <paragraph xml-lang="en-US" id="par_id1039155" role="paragraph" l10n="CHG">The variables for the second variation are calculated as follows:</paragraph> 115 <paragraph xml-lang="en-US" id="par_id7184057" role="code" l10n="NEW">m = EXP(SLOPE(LN(Data_Y);Data_X)) </paragraph> 116 <paragraph xml-lang="en-US" id="par_id786767" role="code" l10n="NEW">b = EXP(INTERCEPT(LN(Data_Y);Data_X)) </paragraph> 117 <paragraph xml-lang="en-US" id="par_id7127292" role="paragraph" l10n="NEW">Calculate the coefficient of determination by</paragraph> 118 <paragraph xml-lang="en-US" id="par_id5437177" role="code" l10n="NEW">r² = RSQ(LN(Data_Y);Data_X) </paragraph> 119 <paragraph xml-lang="en-US" id="par_id6946317" role="paragraph" l10n="NEW">Besides m, b and r² the array function LOGEST provides additional statistics for a regression analysis.</paragraph> 120 <paragraph xml-lang="en-US" id="hd_id6349375" role="heading" level="2" l10n="NEW">The power regression equation</paragraph> 121 <paragraph xml-lang="en-US" id="par_id1857661" role="paragraph" l10n="NEW"> For <emph>power regression</emph> curves a transformation to a linear model takes place. The power regression follows the equation <item type="literal">y=b*x^a</item> , which is transformed to <item type="literal">ln(y)=ln(b)+a*ln(x)</item>.</paragraph> 122 <paragraph xml-lang="en-US" id="par_id8517105" role="code" l10n="NEW">a = SLOPE(LN(Data_Y);LN(Data_X)) </paragraph> 123 <paragraph xml-lang="en-US" id="par_id9827265" role="code" l10n="NEW">b = EXP(INTERCEPT(LN(Data_Y);LN(Data_X)) </paragraph> 124 <paragraph xml-lang="en-US" id="par_id2357249" role="code" l10n="NEW">r² = RSQ(LN(Data_Y);LN(Data_X)) </paragraph> 125 <paragraph xml-lang="en-US" id="hd_id9204077" role="heading" level="2" l10n="NEW">Constraints<comment>UFI: is this still so?</comment></paragraph> 126 <paragraph xml-lang="en-US" id="par_id7393719" role="paragraph" l10n="CHG"> The calculation of the trend line considers only data pairs with the following values:</paragraph> 127 <list type="ordered"> 128 <listitem> 129 <paragraph xml-lang="en-US" id="par_id7212744" role="paragraph" l10n="NEW">logarithm regression: only positive x-values are considered,</paragraph> 130 </listitem> 131 <listitem> 132 <paragraph xml-lang="en-US" id="par_id1664479" role="paragraph" l10n="NEW">exponential regression: only positive y-values are considered,</paragraph> 133 </listitem> 134 <listitem> 135 <paragraph xml-lang="en-US" id="par_id8734702" role="paragraph" l10n="NEW">power regression: only positive x-values and positive y-values are considered.</paragraph> 136 </listitem> 137 </list> 138 <paragraph xml-lang="en-US" id="par_id181279" role="paragraph" l10n="NEW">You should transform your data accordingly; it is best to work on a copy of the original data and transform the copied data.</paragraph> 139 <paragraph xml-lang="en-US" id="hd_id7907040" role="heading" level="2" l10n="NEW">The polynomial regression equation</paragraph> 140 <paragraph xml-lang="en-US" id="par_id8918729" role="paragraph" l10n="NEW">A <emph>polynomial regression</emph> curve cannot be added automatically. You must calculate this curve manually. </paragraph> 141 <paragraph xml-lang="en-US" id="par_id33875" role="paragraph" l10n="NEW">Create a table with the columns x, x², x³, … , xⁿ, y up to the desired degree n. </paragraph> 142 <paragraph xml-lang="en-US" id="par_id8720053" role="paragraph" l10n="NEW">Use the formula <item type="literal">=LINEST(Data_Y,Data_X)</item> with the complete range x to xⁿ (without headings) as Data_X. </paragraph> 143 <paragraph xml-lang="en-US" id="par_id5068514" role="paragraph" l10n="NEW">The first row of the LINEST output contains the coefficients of the regression polynomial, with the coefficient of xⁿ at the leftmost position.</paragraph> 144 <paragraph xml-lang="en-US" id="par_id8202154" role="paragraph" l10n="NEW">The first element of the third row of the LINEST output is the value of r². See the <link href="text/scalc/01/04060107.xhp#Section8">LINEST</link> function for details on proper use and an explanation of the other output parameters.</paragraph> 145 <section id="relatedtopics"> 146 <paragraph xml-lang="en-US" id="par_id4562211" role="paragraph" l10n="CHG"><link href="text/schart/01/04050000.xhp">Y Error Bars tab page</link></paragraph> 147 </section> 148 </body> 149</helpdocument>