Skip to Content

Execute the PAL Auto Smoothing algorithm (Forecast App)

test
0 %
Execute the PAL Auto Smoothing algorithm (Forecast App)
Details

Execute the PAL Auto Smoothing algorithm (Forecast App)

August 13, 2020
Created by
September 6, 2018
Understand and implement the basics of an SAPUI5 application to generate your Forecast results using XSJS services and Machine Learning algorithm in SAP HANA, express edition

You will learn

  • How to use inheritance in controller
  • How use JSON models
  • Create and use Formatters
  • Use the SAPUI5 Router
  • Configure the Application Manifest
  • Create and use Fragments (including Pop Over)
  • How to use an XS OData service (sorting and filtering) in a table and a VizFrame

Step 1: Open the Web IDE

Open the Web IDE, and login using the XSA_DEV credentials.

Switch to the Development tool using the Web IDE Development icon.

Web IDE
Log on to answer question
Step 2: Create JSON Model

In order to drive some of the UI elements, you will define a JSON model. This model will hold the display list for the algorithm and dataset selection.

In the left side panel, expand the forecast/html/resources/webapp/model/algorithms/pal tree node.

Right click on the pal folder node from the tree, and select New > File.

Enter auto_smoothing.json as the file name, then click on OK.

This is the full path of the created file: forecast/html/resources/webapp/model/algorithms/pal/auto_smoothing.json.

Paste the following content:

{
	"key": "auto_smoothing",
	"library": "pal",
	"label": "PAL Auto Smoothing",
	"service": {
		"url": "/xsjs/pal/auto_smoothing.xsjs",
		"method": "POST",
		"params": {
			"ACCURACYMEASURE": {
				"label": "Search strategy",
				"description": "The search strategy for optimal ARMA model",
				"default": 1,
				"values": [{
					"key": "MSE",
					"label": "Mean Square Error"
				}, {
					"key": "MAPE",
					"label": "Mean Absolute Percentage Error"
				}]
			},
			"TRAININGRATIO": {
				"label": "Training Ratio",
				"default": "0.75",
				"step": 0.1,
				"min": 0,
				"max": 1,
				"description": "The ratio of training data to the whole time series. Assuming the size of time series is N, and the training ratio is r, the first N*r time series is used to train, whereas only the latter N*(1-r) one is used to test"
			},
			"SEASONALITYCRITERION": {
				"label": "Seasonal criterion",
				"default": "0.5",
				"step": 0.1,
				"min": 0,
				"max": 1,
				"description": "The criterion of the auto-correlation coefficient for accepting seasonality, in the range of (0, 1). The larger it is, the less probable a time series is regarded to be seasonal."
			},
			"FORECASTNUM": {
				"label": "Number of forecast",
				"default": "20",
				"step": 1,
				"min": 1,
				"max": 100,
				"description": "Number of points to forecast"
			}
		}
	},
	"default_payload": {
		"DATASETNAME": null,
		"ACCURACYMEASURE": "MAPE",
		"TRAININGRATIO": 0.75,
		"SEASONALITYCRITERION": 0.5,
		"FORECASTNUM": 20
	}
}

Save the file using the save icon from the menu.

Log on to answer question
Step 3: Edit the Application Descriptor

Open the manifest.json file located in the forecast/html/resources/webapp folder.

Extend the "models" section with the following element:

"pal_auto_smoothing": {
	"type": "sap.ui.model.json.JSONModel",
	"preload": true,
	"uri": "model/algorithms/pal/auto_smoothing.json"
}

In the "routing" section, extend the "routes" array with the following element:

{
	"name": "auto_smoothing",
	"pattern": "auto_smoothing",
	"target": "auto_smoothing"
}

and extend the "targets" element and add the following element:

"auto_smoothing": {
	"clearAggregation": true,
	"viewName": "algorithms.pal.auto_smoothing"
}

Save the file using the save icon from the menu.

Log on to answer question
Step 4: Create the controller

Expand the forecast/html/resources/webapp/controller/algorithms/pal folder.

Create a new file auto_smoothing.controller.js.

This is the full path of the created file: forecast/html/resources/webapp/controller/algorithms/pal/auto_smoothing.controller.js.

Paste the following content:

sap.ui.define([
	"forecast/html/base/algorithms/Controller"
], function(Controller) {
	"use strict";
	return Controller.extend("forecast.html.controller.algorithms.pal.auto_smoothing", {
		forcedSelectedAlgorithm: "auto_smoothing",
		onPressExecute: function(oEvent) {
			Controller.prototype.onPressExecute.apply(this, oEvent);
				this.setVizProperties("result_fit_viz_frame", "result_fit_popover");			
		}
	});
});

Save the file using the save icon from the menu.

Log on to answer question
Step 5: Create the Fragments

Expand the forecast/html/resources/webapp/fragment/algorithms/pal folder.

Create a new file auto_smoothing_parameters.fragment.xml.

This is the full path of the created file: forecast/html/resources/webapp/fragment/algorithms/pal/auto_smoothing_parameters.fragment.xml.

Paste the following content:

<core:FragmentDefinition xmlns="sap.m" xmlns:core="sap.ui.core" xmlns:ui="sap.ui">
	<ui:layout.form.Form editable="true">
		<ui:layout>
			<ui:layout.form.ResponsiveGridLayout columnsL="1" columnsM="1"/>
		</ui:layout>
		<ui:formContainers>
			<ui:layout.form.FormContainer>
				<ui:formElements>
					<ui:layout.form.FormElement label="{pal_auto_smoothing>/service/params/ACCURACYMEASURE/label}">
						<ui:fields>
							<Select selectedKey="{payload>/ACCURACYMEASURE}" items="{pal_auto_smoothing>/service/params/ACCURACYMEASURE/values}"
								tooltip="{pal_auto_smoothing>/service/params/ACCURACYMEASURE/description}">
								<items>
									<ui:core.ListItem text="{pal_auto_smoothing>label}" key="{pal_auto_smoothing>key}"/>
								</items>
							</Select>
						</ui:fields>
					</ui:layout.form.FormElement>
					<ui:layout.form.FormElement label="{pal_auto_smoothing>/service/params/TRAININGRATIO/label}">
						<ui:fields>
							<Slider progress="true" value="{payload>/TRAININGRATIO}" step="{pal_auto_smoothing>/service/params/TRAININGRATIO/step}"
								min="{pal_auto_smoothing>/service/params/TRAININGRATIO/min}" max="{pal_auto_smoothing>/service/params/TRAININGRATIO/max}"
								enableTickmarks="true" inputsAsTooltips="true" tooltip="{pal_auto_smoothing>/service/params/TRAININGRATIO/description}"/>
							<Input type="Number" value="{payload>/TRAININGRATIO}" enabled="false"/>
						</ui:fields>
					</ui:layout.form.FormElement>
					<ui:layout.form.FormElement label="{pal_auto_smoothing>/service/params/SEASONALITYCRITERION/label}">
						<ui:fields>
							<Slider progress="true" value="{payload>/SEASONALITYCRITERION}" step="{pal_auto_smoothing>/service/params/SEASONALITYCRITERION/step}"
								min="{pal_auto_smoothing>/service/params/SEASONALITYCRITERION/min}" max="{pal_auto_smoothing>/service/params/SEASONALITYCRITERION/max}"
								enableTickmarks="true" inputsAsTooltips="true" tooltip="{SEASONALITYCRITERION>/service/params/SEASONALITYCRITERION/description}"/>
							<Input type="Number" value="{payload>/SEASONALITYCRITERION}" enabled="false"/></ui:fields>
					</ui:layout.form.FormElement>
					<ui:layout.form.FormElement label="{pal_auto_smoothing>/service/params/FORECASTNUM/label}">
						<ui:fields>
							<Slider progress="true" value="{payload>/FORECASTNUM}" step="{pal_auto_smoothing>/service/params/FORECASTNUM/step}"
								min="{pal_auto_smoothing>/service/params/FORECASTNUM/min}" max="{pal_auto_smoothing>/service/params/FORECASTNUM/max}"
								enableTickmarks="true" inputsAsTooltips="true" tooltip="{FORECASTNUM>/service/params/FORECASTNUM/description}"/>
							<Input type="Number" value="{payload>/FORECASTNUM}" enabled="false"/></ui:fields>
					</ui:layout.form.FormElement>
				</ui:formElements>
			</ui:layout.form.FormContainer>
		</ui:formContainers>
	</ui:layout.form.Form>
</core:FragmentDefinition>

Save the file using the save icon from the menu.

Create a new file auto_smoothing_results.fragment.xml.

This is the full path of the created file:

forecast/html/resources/webapp/fragment/algorithms/pal/auto_smoothing_results.fragment.xml

Paste the following content:

<core:FragmentDefinition xmlns="sap.m" xmlns:core="sap.ui.core" xmlns:ui="sap.ui" xmlns:viz="sap.viz">
	<IconTabBar>
		<items>
			<IconTabFilter text="Results">
				<viz:ui5.controls.Popover id="result_popover"/>
				<viz:ui5.controls.VizFrame width="100%" id="result_viz_frame" uiConfig="{applicationSet:'fiori'}" vizType='timeseries_line'>
					<viz:dataset>
						<viz:ui5.data.FlattenedDataset data="{path : 'results>/tables/OUTPUT'}">
							<viz:dimensions>
								<viz:ui5.data.DimensionDefinition name="Date" value="{path : 'results>signal_time'}" dataType="date"/>
							</viz:dimensions>
							<viz:measures>
								<viz:ui5.data.MeasureDefinition name="Original Value" value="{path : 'results>signal_value'}"/>
								<viz:ui5.data.MeasureDefinition name="Forecast" value="{path : 'results>forecast'}"/>
								<viz:ui5.data.MeasureDefinition name="Lower Limit 1" value="{path : 'results>lowerlimit_1'}"/>
								<viz:ui5.data.MeasureDefinition name="Upper Limit 1" value="{path : 'results>upperlimit_1'}"/>
								<viz:ui5.data.MeasureDefinition name="Lower Limit 2" value="{path : 'results>lowerlimit_2'}"/>
								<viz:ui5.data.MeasureDefinition name="Upper Limit 2" value="{path : 'results>upperlimit_2'}"/>
							</viz:measures>
						</viz:ui5.data.FlattenedDataset>
					</viz:dataset>
					<viz:feeds>
						<viz:ui5.controls.common.feeds.FeedItem uid="valueAxis" type="Measure" values="Original Value"/>
						<viz:ui5.controls.common.feeds.FeedItem uid="valueAxis" type="Measure" values="Forecast"/>
				    	<viz:ui5.controls.common.feeds.FeedItem uid="valueAxis" type="Measure" values="Lower Limit 1"/>
						<viz:ui5.controls.common.feeds.FeedItem uid="valueAxis" type="Measure" values="Upper Limit 1"/>
						<viz:ui5.controls.common.feeds.FeedItem uid="valueAxis" type="Measure" values="Lower Limit 2"/>
						<viz:ui5.controls.common.feeds.FeedItem uid="valueAxis" type="Measure" values="Upper Limit 2"/>
						<viz:ui5.controls.common.feeds.FeedItem uid="timeAxis" type="Dimension" values="Date"/>
					</viz:feeds>
				</viz:ui5.controls.VizFrame>
				<ui:table.Table enableBusyIndicator="true" selectionMode="None" width="100%"
					rows="{path : 'results>/tables/OUTPUT', sorter: { path: 'signal_time', descending: true} }">
					<ui:columns>
						<ui:table.Column sortProperty="signal_time" filterProperty="signal_time">
							<Label text="Date"/>
							<ui:template>
								<Text text="{path : 'results>signal_time'}"/>
							</ui:template>
						</ui:table.Column>
						<ui:table.Column sortProperty="signal_value" filterProperty="signal_value">
							<Label text="Original Value"/>
							<ui:template>
								<Text text="{path : 'results>signal_value', formatter : '.formatter.formatNumber'}"/>
							</ui:template>
						</ui:table.Column>
						<ui:table.Column sortProperty="forecast" filterProperty="forecast">
							<Label text="Forecast"/>
							<ui:template>
								<Text text="{path : 'results>forecast', formatter : '.formatter.formatNumber'}"/>
							</ui:template>
						</ui:table.Column>
						<ui:table.Column sortProperty="standard_error" filterProperty="standard_error">
							<Label text="Standard Error"/>
							<ui:template>
								<Text text="{path : 'results>standard_error', formatter : '.formatter.formatNumber'}"/>
							</ui:template>
						</ui:table.Column>
						<ui:table.Column sortProperty="lowerlimit_1" filterProperty="lowerlimit_1">
							<Label text="Lower Limit 1"/>
							<ui:template>
								<Text text="{path : 'results>lowerlimit_1', formatter : '.formatter.formatNumber'}"/>
							</ui:template>
						</ui:table.Column>
						<ui:table.Column sortProperty="upperlimit_1" filterProperty="upperlimit_1">
							<Label text="Upper Limit 1"/>
							<ui:template>
								<Text text="{path : 'results>upperlimit_1', formatter : '.formatter.formatNumber'}"/>
							</ui:template>
						</ui:table.Column>
						<ui:table.Column sortProperty="lowerlimit_2" filterProperty="lowerlimit_2">
							<Label text="Lower Limit 2"/>
							<ui:template>
								<Text text="{path : 'results>lowerlimit_2', formatter : '.formatter.formatNumber'}"/>
							</ui:template>
						</ui:table.Column>
						<ui:table.Column sortProperty="upperlimit_1" filterProperty="upperlimit_2">
							<Label text="Upper Limit 2"/>
							<ui:template>
								<Text text="{path : 'results>upperlimit_2', formatter : '.formatter.formatNumber'}"/>
							</ui:template>
						</ui:table.Column>
					</ui:columns>
				</ui:table.Table>
			</IconTabFilter>
			<IconTabFilter text="Statistics">
				<ui:table.Table enableBusyIndicator="true" selectionMode="None" width="100%" rows="{path : 'results>/tables/STATISTIC'}">
					<ui:columns>
						<ui:table.Column sortProperty="stat_name" filterProperty="stat_name">
							<Label text="Statistics name"/>
							<ui:template>
								<Text text="{path : 'results>stat_name'}"/>
							</ui:template>
						</ui:table.Column>
						<ui:table.Column sortProperty="stat_value" filterProperty="stat_value">
							<Label text="Statistics Value"/>
							<ui:template>
								<Text text="{path : 'results>stat_value'}"/>
							</ui:template>
						</ui:table.Column>
					</ui:columns>
				</ui:table.Table>
			</IconTabFilter>
		</items>
	</IconTabBar>
</core:FragmentDefinition>

Save the file using the save icon from the menu.

Log on to answer question
Step 6: Create the View

Expand the forecast/html/resources/webapp/view/algorithms/pal folder.

Create a new file auto_smoothing.view.xml.

This is the full path of the created file: forecast/html/resources/webapp/view/algorithms/pal/auto_smoothing.view.xml.

Paste the following content:

<mvc:View xmlns:html="http://www.w3.org/1999/xhtml" xmlns:mvc="sap.ui.core.mvc" xmlns="sap.m" xmlns:ui="sap.ui"
 controllerName="forecast.html.controller.algorithms.pal.auto_smoothing"
	displayBlock="true">
	<App id="idAppControl">
		<Page showHeader="true" showNavButton="true" navButtonPress="onNavHome" title="Forecast - PAL Auto Smoothing">
			<content>
				<ui:core.Fragment fragmentName="forecast.html.fragment.display_list" type="XML"/>
				<IconTabBar expandable="false" visible="{config>/enableSelectDataset}" id="tab">
					<items>
						<IconTabFilter key="params" text="Set The Algorithm Execution Parameters">
							<ui:core.Fragment fragmentName="forecast.html.fragment.algorithms.pal.auto_smoothing_parameters" type="XML"/>
							<Button text="Execute" press="onPressExecute"/>
						</IconTabFilter>
						<IconTabFilter key="result" text="Results" visible="{= !!${results>/hasResult} }">
							<ui:core.Fragment fragmentName="forecast.html.fragment.algorithms.pal.auto_smoothing_results" type="XML"/>
						</IconTabFilter>
					</items>
				</IconTabBar>
			</content>
			<footer>
				<Bar>
					<contentLeft>
						<Button text="Back" press="onNavHome"/>
					</contentLeft>
				</Bar>
			</footer>
		</Page>
	</App>
</mvc:View>

Save the file using the save icon from the menu.

Log on to answer question
Step 7: Run the application

Select the html module, then click on the execute icon run from the menu bar.

Once the application is started, the application will open in a new tab/window or you can click on the application URL:

Web IDE

This will open a web page with the following content:

Web IDE
Log on to answer question
Step 8: Validate your results

Select PAL Auto Exponential Smoothing Algorithm, then pick the Ozone level for LA dataset.

Applications

Click on Next.

Leave the parameters with their default values and click on Execute.

Applications

The Results tab will be activated once the algorithm is executed in SAP HANA database.

Applications

Provide an answer to the question below then click on Validate.

In the Statistics tab, what is the value for FORECAST_MODEL_NAME?
×
Step 9: Commit your changes

On the icon bar located on the right side of the Web IDE, click on the Git Pane icon Web IDE.

Click on Stage All, enter a commit comment PAL Auto Smoothing, then click on Commit and Push > origin master.

Log on to answer question

Next Steps

Back to top