Our Thinking
QlikView – Dynamic Data Discovery in 15 Minutes
13 March 2015 | Posted by Terence Sheppard

QlikView_diagramQikView is a popular modern business intelligence (BI) tool that provides a simple and intuitive way to build interactive dashboards that can be shared with others. By searching through these interactive dashboards users can get an understanding of the performance of their organisation and investigate the reasons behind this performance.

Unlike other BI products, QlikView has been designed with the business user in mind, making connecting to data sources, creating visualisations and designing dashboards – simple procedures which require little developer input.

Here I’ll give you an overview of QlikView and show you how it can be used to create useful dashboards and share them online in no time at all.

Connecting to Data Sources

blogDataSourcesOut of the box, QlikView supports loading data from many different sources. These range from flat CSV files to relational databases, all the way up to HDFS via Hive and Imapla ODBC connectors. There is also a wide choice of third party connectors available which can allow you to connect web services such as Google Analytics API and Twitter or NoSQL data stores such as MongoDB. You can even design connectors to your own custom data source if absolutely necessary.

The process of connecting to a data sources is simple and straightforward. You simply point QlikView towards your data source and a wizard will guide you through the field selection and building any clauses in the load statement. At the end of the wizard walkthrough, a loading script is automatically generated and that’s all that QlikView needs.

LoadWizard
Star Schema

When loading data, QlikView creates in-memory associations between fields of different tables based on the field’s name. There is no need to define foreign keys. If two fields are connected just rename them to have the same name in the loading script.

schemaTransform2To get the most out of this in-memory association your data should be organised into a star schema. This schema model is optimised for analytical processing and it greatly improves the performance of the front-end JoinStatementQlikView “apps”. When selections are made on a dashboard, having a star schema means fewer jumps between tables are required to identify associations. A single  table with all transactions and data contained within it is the ideal situation but this may not always be possible.

QlikView makes it easier to perform this transformation into a star schema by providing its own loading and transformation syntax. The advantage is that data from disparate sources such as flat files and various databases can easily be integrated into a single data model.

Creating Visualisations

createDash1After creating the data model it’s time to create visualisations. It is these visualisations which are combined with data selection tables to form the interactive dashboards. Creating tables, graphs and charts is again a simple and intuitive process with QlikView providing a wizard to guide you. But don’t let this simplicity fool you, as QlikView has many advanced features. Its use of expression and set analysis mean that you can compare any subset of the total data to any other subset to create incredibly useful visualisations. Features such as drill-through and drill-down are easily set up within the visualisations and data can be downloaded from the dashboard if required.dashboard1

Online Viewing, Reloading and Scheduling

QVmobileOnce your dashboards have been created they can be easily shared online with others using the QlikView Server. Just transfer the QlikView app file to the specified folder on the QlikView server installation. The app can then be accessed via a dedicated Windows desktop application or via a web portal. The web portal is the more exciting option as it can be accessed via a web browser on a range of platforms including mobile.

accessPointYou can also set your dashboards to reload from the data sources automatically on a schedule. This means your dashboards will always contain the most up-to-date information available. You can also schedule static reports, created using visualisations from our dashboards, to be distributed as PDF files to a mailing list. Features like reduce and distribute mean that reports can be changed to show only information relating to a users field of interest.

Wrap-Up

dashboard3There you have it, a quick run through of some of the possibilities of QlikView. Although there are a plethora of features I’ve not covered, you now have all the basic information you need to get a QlikView dashboard up and running. The most difficult part now is to decide what exactly you and others within your organisation need to know, but as QlikView puts these decisions in the hands of the person best able to answer this, you the business user, this is not much of a hurdle at all.

No comments
Leave a Comment