Add Trend Lines to Presentation Charts in C++
Overview
This article explains how to add trend lines to presentation charts by using Aspose.Slides. It shows how to create a chart, add trend lines to chart series, and work with several trend line types, including exponential, linear, logarithmic, moving average, polynomial, and power.
It also describes how to add a custom line to a chart by inserting a line shape, and includes a short FAQ about forward and backward trend line projection values and whether trend lines are preserved during export to PDF or SVG and when rendering charts as images.
Add a Trend Line
Use ITrendlineCollection::Add to add a trend line to a chart series, selecting its type with TrendlineType. The example creates a clustered column chart with default data and adds two trend lines to each of its three series.
The exponential trend line hides its equation and R-squared value through ITrendline::set_DisplayEquation and ITrendline::set_DisplayRSquaredValue, while the linear trend line is red. The logarithmic trend line uses IOverridableText::AddTextFrameForOverriding for a custom label. The moving average uses ITrendline::set_Period to average three data points and ITrendline::set_TrendlineName for a custom name.
The polynomial trend line uses ITrendline::set_Order to set its degree to three and extends one category forward through ITrendline::set_Forward. The power trend line extends one category backward through ITrendline::set_Backward.
#include <system/shared_ptr.h>
#include <drawing/color.h>
#include <DOM/Presentation.h>
#include <DOM/ISlide.h>
#include <DOM/IShapeCollection.h>
#include <DOM/ILineFormat.h>
#include <DOM/IFillFormat.h>
#include <DOM/IColorFormat.h>
#include <DOM/FillType.h>
#include <DOM/IChart.h>
#include <DOM/Chart/ChartType.h>
#include <DOM/Chart/IChartData.h>
#include <DOM/Chart/IChartSeries.h>
#include <DOM/Chart/IChartSeriesCollection.h>
#include <DOM/Chart/ITrendline.h>
#include <DOM/Chart/ITrendlineCollection.h>
#include <DOM/Chart/TrendlineType.h>
#include <DOM/Chart/IFormat.h>
#include <Export/SaveFormat.h>
using namespace System;
using namespace System::Drawing;
using namespace Aspose::Slides;
using namespace Aspose::Slides::Charts;
using namespace Aspose::Slides::Export;
auto presentation = MakeObject<Presentation>();
auto slide = presentation->get_Slide(0);
auto chart = slide->get_Shapes()->AddChart(ChartType::ClusteredColumn, 20, 20, 500, 400);
auto firstSeries = chart->get_ChartData()->get_Series()->idx_get(0);
auto exponentialTrendline = firstSeries->get_TrendLines()->Add(TrendlineType::Exponential);
exponentialTrendline->set_DisplayEquation(false);
exponentialTrendline->set_DisplayRSquaredValue(false);
auto linearTrendline = firstSeries->get_TrendLines()->Add(TrendlineType::Linear);
linearTrendline->get_Format()->get_Line()->get_FillFormat()->set_FillType(FillType::Solid);
linearTrendline->get_Format()->get_Line()->get_FillFormat()->get_SolidFillColor()->set_Color(Color::get_Red());
auto secondSeries = chart->get_ChartData()->get_Series()->idx_get(1);
auto logarithmicTrendline = secondSeries->get_TrendLines()->Add(TrendlineType::Logarithmic);
logarithmicTrendline->AddTextFrameForOverriding(u"New log trend line");
auto movingAverageTrendline = secondSeries->get_TrendLines()->Add(TrendlineType::MovingAverage);
movingAverageTrendline->set_Period(3);
movingAverageTrendline->set_TrendlineName(u"New TrendLine Name");
auto thirdSeries = chart->get_ChartData()->get_Series()->idx_get(2);
auto polynomialTrendline = thirdSeries->get_TrendLines()->Add(TrendlineType::Polynomial);
polynomialTrendline->set_Forward(1);
polynomialTrendline->set_Order(3);
auto powerTrendline = thirdSeries->get_TrendLines()->Add(TrendlineType::Power);
powerTrendline->set_Backward(1);
presentation->Save(u"ChartTrendLines.pptx", SaveFormat::Pptx);
Set a Trend Line Intercept
Use ITrendline::set_Intercept to fix a trend line’s Y value at X = 0. This setting is supported for linear, exponential, and polynomial trend lines.
For example, a production cost model may need to start at known fixed costs of 100, even when measurements at nonzero production volumes suggest a different baseline. For a linear trend line, fixing the intercept constrains the equation y = ax + b to b = 100; the slope is fitted to the data under that constraint. It does not simply move an already fitted line. Axis crossing settings control where the chart axes meet, not the coefficients of the fitted trend line.
The following example creates a scatter (XY) chart with production volumes from 1 to 4 and costs of 170, 200, 230, and 260. It saves the chart with the default automatic intercept, sets the intercept to 100, and saves a second PPTX. The two files differ only in the intercept setting. The trend line displays its equation and extends back to X = 0.
#include <system/shared_ptr.h>
#include <system/object_ext.h>
#include <DOM/Chart/IChartDataWorkbook.h>
#include <DOM/Chart/IChartDataCell.h>
#include <DOM/Chart/IChartCategoryCollection.h>
#include <DOM/Chart/IChartDataPointCollection.h>
#include <DOM/Presentation.h>
#include <DOM/ISlide.h>
#include <DOM/IShapeCollection.h>
#include <DOM/IChart.h>
#include <DOM/Chart/ChartType.h>
#include <DOM/Chart/IChartData.h>
#include <DOM/Chart/IChartSeries.h>
#include <DOM/Chart/IChartSeriesCollection.h>
#include <DOM/Chart/ITrendline.h>
#include <DOM/Chart/ITrendlineCollection.h>
#include <DOM/Chart/TrendlineType.h>
#include <Export/SaveFormat.h>
using namespace System;
using namespace Aspose::Slides;
using namespace Aspose::Slides::Charts;
using namespace Aspose::Slides::Export;
auto presentation = MakeObject<Presentation>();
auto slide = presentation->get_Slide(0);
auto chart = slide->get_Shapes()->AddChart(ChartType::ScatterWithMarkers, 20, 20, 480, 210);
chart->set_HasLegend(false);
chart->get_ChartData()->get_Series()->Clear();
chart->get_ChartData()->get_Categories()->Clear();
auto workbook = chart->get_ChartData()->get_ChartDataWorkbook();
workbook->Clear(0);
auto nameCell = workbook->GetCell(0, 0, 1, ObjectExt::Box<String>(u"Production cost"));
auto series = chart->get_ChartData()->get_Series()->Add(nameCell, ChartType::ScatterWithMarkers);
const double costs[] = { 170, 200, 230, 260 };
for (auto i = 0; i < 4; i++)
{
auto xCell = workbook->GetCell(0, i + 1, 0, ObjectExt::Box<int32_t>(i + 1));
auto yCell = workbook->GetCell(0, i + 1, 1, ObjectExt::Box<double>(costs[i]));
series->get_DataPoints()->AddDataPointForScatterSeries(xCell, yCell);
}
auto trendline = series->get_TrendLines()->Add(TrendlineType::Linear);
trendline->set_DisplayEquation(true);
trendline->set_Backward(1);
presentation->Save(u"AutomaticIntercept.pptx", SaveFormat::Pptx);
trendline->set_Intercept(100);
presentation->Save(u"FixedIntercept.pptx", SaveFormat::Pptx);
With automatic calculation, these data produce y = 30x + 140. Setting the intercept to 100 instead produces approximately y = 43.33x + 100: the fitted value at zero matches the known baseline, while the line no longer passes through every data point.
| Automatic intercept | Fixed intercept of 100 |
|---|---|
 |
 |
Both images use the same data and axis scales. The trend lines are extended back to X = 0 to show their intercepts.
By default, the property contains std::numeric_limits<double>::quiet_NaN(), which represents an unspecified intercept rather than zero. The intercept is then calculated automatically from the data.
Add a Custom Line
A custom line is a shape drawn over the chart rather than a trend line calculated from series data. This example uses IShapeCollection::AddAutoShape to add a horizontal line to the slide over the chart. The line’s position is specified in slide coordinates, measured in points, so it starts at the chart’s left edge and sits halfway down the chart. Its width matches the chart’s width. The line format returned by IShape::get_LineFormat gives the line a solid red color.
#include <system/shared_ptr.h>
#include <drawing/color.h>
#include <DOM/Presentation.h>
#include <DOM/ISlide.h>
#include <DOM/IShapeCollection.h>
#include <DOM/IAutoShape.h>
#include <DOM/ShapeType.h>
#include <DOM/ILineFormat.h>
#include <DOM/IFillFormat.h>
#include <DOM/IColorFormat.h>
#include <DOM/FillType.h>
#include <DOM/IChart.h>
#include <DOM/Chart/ChartType.h>
#include <Export/SaveFormat.h>
using namespace System;
using namespace System::Drawing;
using namespace Aspose::Slides;
using namespace Aspose::Slides::Charts;
using namespace Aspose::Slides::Export;
auto presentation = MakeObject<Presentation>();
auto slide = presentation->get_Slide(0);
auto chart = slide->get_Shapes()->AddChart(ChartType::ClusteredColumn, 100, 100, 500, 400);
auto line = slide->get_Shapes()->AddAutoShape(ShapeType::Line, chart->get_X(), chart->get_Y() + chart->get_Height() / 2, chart->get_Width(), 0);
line->get_LineFormat()->get_FillFormat()->set_FillType(FillType::Solid);
line->get_LineFormat()->get_FillFormat()->get_SolidFillColor()->set_Color(Color::get_Red());
presentation->Save(u"AddCustomLines.pptx", SaveFormat::Pptx);
FAQ
What do ‘forward’ and ‘backward’ mean for a trend line?
They are the lengths of the trend line projected forward/backward: for scatter (XY) charts — in axis units; for non-scatter charts — in number of categories. Only non-negative values are allowed.
Will the trend line be preserved when exporting the presentation to PDF or SVG, or when rendering a slide to an image?
Yes. Aspose.Slides converts presentations to PDF or SVG and renders charts to images; trend lines, as part of the chart, are preserved during these operations. A method is also available to export an image of the chart itself.