Predictive Charts

Predictive charts let you visualize current data along with predicted/forecasted data within a given value range.

Figure 1: Predicted/forecasted data is surrounded by a highlighted area of possible values.

This tutorial shows how to configure your portlet to use predictive charts. Follow these steps:

Import the chart taglib along with the PredictiveChartConfig and MixedDataColumn classes into your bundle’s init.jsp file:

<%@ taglib prefix="chart" uri="http://liferay.com/tld/chart" %>
<%@ page import="com.liferay.frontend.taglib.chart.model.predictive.PredictiveChartConfig" %>
<%@ page import="com.liferay.frontend.taglib.chart.model.MixedDataColumn" %>

Add the following Java scriptlet to the top of your view.jsp. Add a MixedDataColumn object —a column that supports both single number values and arrays of three numbers—for each data series. Single number values define existing data. Arrays of numbers are used as the prediction/forecast data and contain three numbers: a minimum value, an estimated value, and a maximum value. The estimated value is rendered solid and surrounded by a highlighted area with borders specified by the minimum and maximum values. This lets you visualize your estimated values, while also giving you an idea of the possible value ranges. Use the addDataColumn() method to add each data series:
```
<%
private PredictiveChartConfig _predictiveChartConfig = new
PredictiveChartConfig();

MixedDataColumn mixedDataColumn1 = new MixedDataColumn(
  "data1", 130, 340, 200, 500, 80, 240, 40,
  new Number[] {370, 400, 450}, new Number[] {210, 240, 270},
  new Number[] {150, 180, 210}, new Number[] {60, 90, 120},
  new Number[] {310, 340, 370}
);

_predictiveChartConfig.addDataColumn(mixedDataColumn1);

MixedDataColumn mixedDataColumn2 = new MixedDataColumn(
  "data2", 210, 160, 50, 125, 230, 110, 90,
  Arrays.asList(170, 200, 230), Arrays.asList(10, 40, 70),
  Arrays.asList(350, 380, 410), Arrays.asList(260, 290, 320),
  Arrays.asList(30, 70, 150)
);

_predictiveChartConfig.addDataColumn(mixedDataColumn2);

_predictiveChartConfig.setAxisXTickFormat("%b");

_predictiveChartConfig.setPredictionDate("2018-07-01");

List<String> timeseries = new ArrayList<>();

timeseries.add("2018-01-01");
timeseries.add("2018-02-01");
timeseries.add("2018-03-01");
timeseries.add("2018-04-01");
timeseries.add("2018-05-01");
timeseries.add("2018-06-01");
timeseries.add("2018-07-01");
timeseries.add("2018-08-01");
timeseries.add("2018-09-01");
timeseries.add("2018-10-01");
timeseries.add("2018-11-01");
timeseries.add("2018-12-01");

_predictiveChartConfig.setTimeseries(timeseries);
%>
```
Predictive charts have the following properties:

axisXTickFormat: An optional string which specfies the time formatting on the X axis. For more information on which formats can be specified please refer to d3’s time format README. This value is set using the setAxisXTickFormat() method.

Prediction Date: A date as a string that represents the point in the timeline from when the forecast/prediction is shown. This value is parsed as a Date object in JavaScript and set using the setPredictionDate() method.

Time Series: A timeline for the data which is displayed on the X axis of the chart. This value is set as an array of dates (2018-01-01 for example).
Add the <chart> taglib to the view.jsp, passing the _predictiveChartConfig as the config attribute’s value:
```
<chart:predictive
  config="<%= _predictiveChartConfig %>"
/>
```