This tutorial is part of a series that explores telecom call drop predictions using IBM Cloud Pak for Data, data virtualization, Watson OpenScale, and Cognos Analytics.
|201||Query across distributed data sources as one: Data virtualization for data analytics||Tutorial|
|201||Monitor your machine learning models using Watson OpenScale in IBM Cloud Pak for Data||Pattern|
|301||Build dashboards in Cognos Analytics on IBM Cloud Pak for Data||Tutorial|
|301||Predict, manage, and monitor the call drops of cell towers using IBM Cloud Pak for Data||Pattern|
IBM Cognos Analytics on IBM Cloud Pak for Data integrates reporting, modeling, analysis, and exploration so you can understand your organization’s data – and make effective business decisions. It’s a proven, self-service, dependable analytics solution that provides scalability and analytics governance. IBM Cognos Analytics can help you create visualizations and dashboards without needing a data science background.
The IBM Cognos Analytics add-on in IBM Cloud Pak for Data helps make it easier for you to extract insights from your underlined data with features, including:
- Automated data preparation
- Automated modeling
- Automated visualizations and dashboards
- Data exploration
The add-on also helps enable you to share your findings with others using visualizations, dashboards, reports, and interactive stories.
In this tutorial, we’ll demonstrate how to create a dashboard using IBM Cognos Analytics on IBM Cloud Pak for Data.
Scenario considered for this tutorial:
In this tutorial, you’re a Telecom servicer provider who is responsible to figure out why call drops are frequently happening for any selected towers. You need to understand the reasons for the call drops–as well as the contributing factors influencing the call drops for the selected tower. You’re getting these insights from the Python model output. Using these details and the IBM Cognos Analytics dashboard feature, you should be able to build a dashboard that depicts call drop prediction for the next 24 hours.
After completing this tutorial, you’ll learn how to:
- Build an IBM Cognos Analytics dashboard on IBM Cloud Pak for Data.
- Launch IBM Cognos Analytics on IBM Cloud Pak for Data.
- Import and export the IBM Cognos Analytics dashboard binaries in the IBM Cloud Pak for Data environment.
One or more csv files as sources. Note: For this tutorial, we’ll demonstrate with:
Completing this tutorial should take about 30 minutes.
Step 1: Launch an IBM Cloud Pak for Data instance
- To launch IBM Cloud Pak for Data instance, use the instance URL
CP4D. Sample URL as follows: ‘https://IP:port/zen/#/gettingStarted‘
- After you log in with your credentials, you should see the welcome page below.
- Ensure the IBM Cloud Pak for Data version is 18.104.22.168 or greater.
Step 2: Open the IBM Cognos Analytics add-on in IBM Cloud Pak for Data
To open Cognos Analytics from IBM Cloud Pak for Data, complete the following steps:
- On the welcome screen, click on the add-on icon.
- Click on the navigation menu. Under the Analytics category, you should see Cognos Analytics as enabled. If it is NOT enabled, here is the link to Enable IBM Cognos Analytics on IBM Cloud Pak for Data. This tool is available only if your administrator configures the Cognos Analytics add-on.
- Once you have selected the Cognos Analytics, click on the browse icon and open Cognos Analytics.
- Click on Cognos Analytics under
- The Cognos Analytics welcome screen will look like screen below.
Step 3: Connecting to Db2 (or) Uploading .csv file to Cognos Analytics for building dashboards
- Let’s start by getting data for the tutorial.
Connecting to Db2 db, which is part of the IBM Cloud Pak for Data cluster for building dashboards
- From the Cognos Analytics welcome screen, go to Manage and click on ‘Data Server Connections’ to create a new connection to the Db2 db.
- Now, click on Data server ‘+’ icon for the new Data Server connection.
Select a type as
IBM Db2and then edit the JDBC details and add the details for the following:
- HostName (Ip of the database server which is part of the IBM Cloud Pak for Data cluster)
Authentication method, select the radio button for
Prompt the user ID and passwordand click on
- Now, it will prompt for database credentials as shown in below screenshot. Provide the database credentials. You should get these credentials if you follow the Code pattern.
- Provide the Db2 connection name as per your naming convention. Upon successful Connection, message to the database and click on save.
- In the same window, under the Schemas tab, select the schema name where your data is actually loaded. Next, click on
Load metadata. In this case, we loaded the data in
DB2INST1schema. All the database tables under that schema, the meta data of the tables will be loaded to Cognos Analytics content store.
Follow the steps below to create a data module to make the data source available for Cognos Analytics dashboard
Data Modulefrom the Cognos Analytics welcome page screen by clicking the
+icon from left bottom screen. See below screenshot.
Select Sourceswindow click on
Data Servers and Schemastab.
- Now select the
Db2_connectionas a source (which was created based on the previous steps).
- Select the
DB2INST1schema and click OK.
- Now add tables to the data module that you want to use for building your dashboard. Click on the
Select tablesicon and click
- Select the table
Times Series Dataand click
Refreshbutton, and click ok.
- A new unsaved data model gets created, click on
- Next, create a new folder by name
Data Moduleunder the Team Content tab.
- Under the Data Module folder, save the data module as
- Once you’ve completed creating the Data module, go to Step 4 (in creating a dashboard).
Note: Remember that when adding data source from the dashboard, you'll have to select the source `TimeSeriesDataModule` which is under the `Data Module` folder of `Team Content`.
Uploading .csv file to Cognos Analytics for building dashboard
Go to the data folder
data/Telco_Time_Series_DataSet.csvand download the file to your local machine.
You can also click the
Newicon and then click
- Go to where you saved
Telco_Time_Series_DataSet.csvand select it.
After successful upload of the file, you should same screen below.
- Check whether the data asset appears in the My content folder.
Step 4: Creating a dashboard
- On the home page, click the
Newicon at the bottom of the window and click on
- You’ll see predefined templates for easy arrangement and alignment of visualizations and other elements in a dashboard. Select the blank template and then click OK.
Now add a data asset to this dashboard so that we can build some insights through visualizations.
- In the
Datapanel, click the
Add a sourceicon.
Go to the My content folder and select Telco_Time_Series_DataSet.csv. Click Add.
Expand the data asset to see the columns available.
- In the
Creating Latitude and Longitude based Map:
Near by Towers - Call Drop Analysis
longcolumns from the Data Source and Drag and drop to the left top pane.
Expand the chart to add additional data items.
Call Droppedfields under Latitude/Longitude area under label, size and color.
Next, minimize the chart.
Select the Map chart and click on properties icon to customize the chart properties.
Set the properties for the chart as shown below for better visualization.
Generaltab, enable the title and add appropriate title to the chart.
Creating Time Series Line Chart:
24 Hour Call Drop Prediction
- Select the Charts icon, drag and drop to the dashboard pane.
Start_Timeto x-axis and
Predicted_Call_Drop_Percto y-axis as seen in the below screenshot.
- Now select the Line chart and click on properties icon to customize the chart for better visuals.
Creating the Bar chart:
Historical Trends - Call Drop by Weather & Traffic(Co-Relation)
- Select the Bar chart icon, drag and drop to the dashboard pane area. See below screenshot.
trafficcolumn to bars field and
Predicted_Call_Drop_Percto #Length filed as seen in the below screenshot.
If the user is trying to upload incorrect files other than csv files, you would see the below error.
To Resolve the above error, ensure you upload the csv file ONLY.
In this tutorial, you learned how to create a dashboard using IBM Cognos Analytics on IBM Cloud Pak for Data. The typical flow of a data and analytics project usually begins with understanding the data, building the data model, building AI or data science models, and building visualizations (for the end user) to extract the insights from the data. This tutorial showed you how IBM Cognos Analytics in IBM Cloud Pak for Data can help you to seamlessly build dashboards on the fly.