[labplot/analysis_playground] /: added documentation for numerical integration

[Stefan Gerlach]

Git commit c626c798c69132686f01e420d94d34aea5eb90d5 by Stefan Gerlach.
Committed on 30/09/2016 at 15:59.
Pushed by sgerlach into branch 'analysis_playground'.

added documentation for numerical integration

M  +33   -10   doc/index.docbook
M  +2    -2    src/backend/nsl/nsl_int.c

http://commits.kde.org/labplot/c626c798c69132686f01e420d94d34aea5eb90d5

diff --git a/doc/index.docbook b/doc/index.docbook
index 892ac8f..352b2bd 100644
--- a/doc/index.docbook
+++ b/doc/index.docbook
@@ -51,8 +51,8 @@
 </copyright>
 
 <legalnotice>&FDLNotice;</legalnotice>
-<date>2016-09-26</date>
-<releaseinfo>3.2.5</releaseinfo>
+<date>2016-09-30</date>
+<releaseinfo>3.2.6</releaseinfo>
 
 <abstract>
 	<para>
@@ -89,7 +89,7 @@ Features:
 <listitem><para>Cartesian plots, created either from imported or manually created data sets or via mathematical equation</para></listitem>
 <listitem><para>Definition of mathematical formulas is supported by syntax-highlighting and completion and by the list of thematicaly grouped mathematical and physical constants and functions</para></listitem>
 <listitem><para>Investigation of plotted data is supported by many zooming and navigation features</para></listitem>
-<listitem><para>Several analysis functions and methods for data reduction, differentiation, interpolation, smoothing, (nonlinear) fitting, Fourier filter and Fourier transform</para></listitem>
+<listitem><para>Several analysis functions and methods for data reduction, differentiation, integration, interpolation, smoothing, (nonlinear) fitting, Fourier filter and Fourier transform</para></listitem>
 <listitem><para>Linear and non-linear fits to data, several fit-models are predefined and custom models with arbitrary number of parameters can be provided</para></listitem>
 <listitem><para>Datapicker for manual or (semi-)automatic data extraction from imported images containing plots and curves.</para></listitem>
 </itemizedlist>
@@ -792,11 +792,15 @@ The menu is only available when a datapicker object is selected on the <guilabel
 
 <chapter id="analysis">
   <title>Analysis functions</title>
-  <para>&LabPlot; supports a wide variety of data analysis functions:
+  <sect1 id="analysis_overview">
+    <title>Overview</title>
+  <para>
+    &LabPlot; supports a wide variety of data analysis functions:
     </para>
       <itemizedlist>
 	<listitem><para>Data reduction</para></listitem>
 	<listitem><para>Differentiation</para></listitem>
+	<listitem><para>Integration</para></listitem>
 	<listitem><para>Interpolation</para></listitem>
 	<listitem><para>Smoothing</para></listitem>
 	<listitem><para>Nonlinear curve fitting</para></listitem>
@@ -806,7 +810,9 @@ The menu is only available when a datapicker object is selected on the <guilabel
     <para>
     All of them can be applied to any data consisting of x- and y-columns. 
     The analysis functions can be accessed using the Analysis menu or the context menu of a worksheet.
-    The newly created curves can be customized (line style, symbol style, &etc;) like any other x-y-curve.</para>  
+    The newly created curves can be customized (line style, symbol style, &etc;) like any other x-y-curve.
+    </para>
+    </sect1>
 
   <sect1 id="data_reduction">
     <title>Data reduction</title>
@@ -841,7 +847,24 @@ The menu is only available when a datapicker object is selected on the <guilabel
 	<listitem><para>order of accurary (up to 4th order, depending on derivation order)</para></listitem>
       </itemizedlist>
   </sect1>
-    
+  
+  <sect1 id="integration">
+    <title>Integration</title>
+    <para>
+      Numerical integration of data can be done specifying one of the methods
+    </para>
+      <itemizedlist>
+	<listitem><para>rectangle (1-point) rule</para></listitem>
+	<listitem><para>trapezoid (2-point) rule</para></listitem>
+	<listitem><para>Simpson-1/3 (3-point) rule</para></listitem>
+	<listitem><para>Simpson-3/8 (4-point) rule</para></listitem>
+      </itemizedlist>
+   <para>
+     The default method (trapezoid) should be suitable for most cases. 
+     The number of resulting data points is reduced for both Simpson-rules due to the properties of these methods.
+   </para>
+  </sect1> 
+
   <sect1 id="interpolation">
     <title>Interpolation</title>
     <para>
@@ -936,10 +959,10 @@ The menu is only available when a datapicker object is selected on the <guilabel
       Following options can be used to suite one needs:
     </para>
       <itemizedlist>
-	<listitem><para>window function (Welch, Hann, Hamming, etc.) to avoid leakage effects</para></listitem>
-	<listitem><para>output (magnitude, amplitude, phase, dB, etc.)</para></listitem>
-	<listitem><para>one or two sided spectrum with or without shifting</para></listitem>
-	<listitem><para>x axis scaling to frequency, index or period</para></listitem>
+	<listitem><para>Window function (Welch, Hann, Hamming, etc.) to avoid leakage effects</para></listitem>
+	<listitem><para>Output (magnitude, amplitude, phase, dB, etc.)</para></listitem>
+	<listitem><para>One or two sided spectrum with or without shifting</para></listitem>
+	<listitem><para>X axis scaling to frequency, index or period</para></listitem>
       </itemizedlist>
   </sect1>
 </chapter>
diff --git a/src/backend/nsl/nsl_int.c b/src/backend/nsl/nsl_int.c
index ce27299..d5adf5a 100644
--- a/src/backend/nsl/nsl_int.c
+++ b/src/backend/nsl/nsl_int.c
@@ -26,8 +26,8 @@
  ***************************************************************************/
 
 /* TODO:
-	* absolute area for Simpson/Simpson-3/8 rules?
-	* consider not multiple of 3/4 points for Simpson/Simpson-3/8 rules?
+	* absolute area for Simpson/Simpson-3/8 rules (needs more numerics)
+	* start and end point for integration: use ranges?
 */
 
 #include <stdio.h>