What is Tableau?
Data Visualization binds or maps data to the perceivable and consistent representations to gather a quick and holistic understanding about various aspects of data like patterns, distributions, trends, and anomalies in the massive data corpus.
The inherent abilities of visualization tools like the ease of usage, enhanced interaction with data elements, the extent of generalization towards high-dimensional data, etc. made data visualization to evolve as a key discipline in the data-driven world.
Need, Scope and Features
The overwhelming growth rate of available data volumes and pressing need for the decision-makers in all domains of business and research to make quick and accurate decisions increased the scope for data visualization tools to understand data with graphics. The requisite efforts and time to create, maintain packages of the graphics and level of expertise to understand graphics jargons is greatly reduced.
The Salient features offered by Data Visualization tools are:
- Ease use of tool functionalities avoids the need for in-depth training on technology for creation or interaction of views in order to derive valuable insights.
- Scalability to the rapidly growing data size and to the increased data complexity comprising scalar, vector, and tensor data types without compromising the performance of the tool.
- Ability to handle heterogeneous data by possessing data-connectors or parsers for various sources that hold or store
- Structural data as in Relational Databases (Oracle, MySQL, PostgreSQL)
- Semi-structured data like XML files, JSON payloads
- Unstructured data in NoSQL databases like MongoDB, Couchbase, etc.
- Cloud data storage like Oracle Cloud, AWS and Microsoft Azure, etc.
- Integration of view manipulation features like slicing and dicing, roll up to the required level of detail, filtering the data to help in detailed explorative data analysis
- Support for semantic visualization to perform an automatic selection of suitable visualization primitives such as line-charts, bar-charts, scatter-plots, Tree diagram, etc. for nominal, ordinal and numerical data.
One of the most popular data visualization tools is Tableau which proved to be the best in visual-based data exploration.
Working with Tableau
The brief description of key concepts to work with Tableau is:
-
Dimensions and Measures
Tableau classifies the data into Dimensions and Measures. The independent fields that cannot be aggregated considered as Dimensions and those fields that can be aggregated and are context-dependent are referred to as Measures
For example, consider Superstore dataset in which Region, Year are dimensions that provide context to add meaning to the Sales figure which is a summarizable measure
-
Level of Detail
The three expressions supported by Tableau to traverse levels of detail are
1. INCLUDE – Aggregation is performed post grouping the INCLUDED list of dimension fields
{ INCLUDE [Field 1],[Field 2]: SUM([column]) }
2. EXCLUDE – Aggregation is performed post grouping the dimension fields that are NOT present in the EXCLUSION list
4.5 (2,366 ratings)
View Course
{EXCLUDE [Field1], [Field2], [Field3] : sum([column])}
3. FIXED – Unlike Include and Exclude expressions, FIXED aggregated values are insensitive i.e. remain unchanged to the altered order of displayed dimensions (addition of new dimension or hiding existing dimension) in the view.
{ FIXED [Field 1]:SUM([column])}
-
Parameters
Visualization interaction can be enhanced by parameterizing a few aspects of the view. The common use-cases that need usage of parameters are
- The inclusion of Top-N Filters
- Bin Size adjustment of histograms
- Reference line definition for threshold-based classification
The range or list of parameter values, users can change or choose, should be of the same data type as defined in parameter properties
Generally, the process for the parameters used in Tableau involves below steps:
1. Create New Parameter: Every instance of entering a constant during view creation supports the option – ‘create new parameter’ to enable tuning of the concerned aspect in published view
2. Show Parameter Control: Parameter controls like Filter boxes, gauges, sliding bars, etc. can be made visible in the view by selecting the ‘show parameter control’ by right-clicking against the parameters in parameter pane.
3. Use Parameter in Calculations: Parameter values returned from the parameter control can be used in calculations, functions or in logical statements so as to provide a provision to users for depicting desired measure/statistic.
-
Sets
Sets are manually-created or conditionally computed subset of input data to facilitate customized depiction elements in the view. The features of the sets are:
1. Dynamic groups: Unlike categories due to a grouping which are static and can be created within the dimensions having a large membership, computed sets based on the conditional logic defined on either dimensions or measures dynamically categorizes fields whenever underlying data changes.
2. Reusability: Though filters support conditional categorization of dimensions or measures, the scope of filters is limited to that dimension in workbook whereas sets can be saved and reused in calculations or as the component in another set.
3. Order of precedence: Sets have higher precedence than dimension filters or measure filters in the query pipeline so that the list of values displayed to choose in the filters of view will be optimum as per set definition.
-
Table Calculations
Table Calculations operates on data locally visible in the Tableau view to derive measures like running totals within sub-categories, cumulative sum, moving average with tunable window length, the ranking of field values, percent differences, etc.
Tableau Products
Various Tableau products are categorized into:
- Visualization Development products that include Tableau Desktop and Tableau Public (Desktop)
- Visualization Publishing products that include Tableau Server, Tableau Reader, Tableau Online, Tableau Public(Server)
-
Tableau Desktop
Tableau Desktop allows users to create, format and integrate various interactive views and dashboards using the rich set of visualization primitives. It also supports live-up to-date analysis by querying data residing in various native and live-connected databases.
The created visualizations are published by sharing the Tableau Packaged Workbook with extension .twbx that comprises of
- Tableau Workbook with extension .twb which is an XML document describing visualization templates
- Tableau Data Extracts with extension .tde file which is compressed data source file
- Other supporting files like images etc.
-
Tableau Server
Tableau server is reliable, secure and well-governed enterprise level environment to share and publish visualizations created using Tableau Desktop. This product acts as a central repository for various data sources in the data engine, users’ security roles and access privilege details and all visualizations across the entire firm.
-
Tableau Public
Tableau Public is a cloud-hosted free version with tool-usage limitations and is bundled with two sub products namely Tableau Public (Desktop) and Tableau Public (Server). The views are created in Tableau Public (Desktop) which are saved and published in Tableau Public(server).
The notable limitations with Tableau Public are:
- Locally available data extracts are the only data sources supported
- The maximum limit on the number of rows to be inputted is one million
- Unlike Tableau Desktop, users cannot save the report locally and are restricted to save the workbook in Tableau Public(Server) which is accessible to all of its user
-
Tableau Online
Tableau Online is a cloud-hosted view sharing platform with an ability to connect with cloud databases like Amazon Redshift, Google BigQuery, etc. Scheduled refresh of extracts and live-connection with on-premises datastores are done using tableau bridge.
- The inherent benefit offered by the cloud bases solution is reduced infrastructure cost for its users by avoiding associated upgrades, patching and maintenance activities
- Unlike Tableau server, editing of workbooks and visualizations need data server connection and these operations are limited by maximum bound on row count.
-
Tableau Reader
Tableau Reader is a desktop application that enables users to open and perform view interactions like drill-down and roll-up of OLAP cubes, filtering on dimensions, etc. but cannot edit the embedded content in the published visualizations which are built in Tableau Desktop.
Conclusion
Tableau gained positive customer experience and positioned as a leader in Business Intelligence (BI) vendor space. The remarkable capability of translating raw data to visual representations was achieved by offering a large set of data connectors and an easily accessible rich set of visualizations primitives.
Amid changing customer needs and growing user expectations, Tableau is aiming to support machine-learning enabled data preparation and smart data discovery. Alongside, Tableau is focusing to address challenging gaps in enterprise features which are listed below:
- Comprised performance while integrating large and complex data from heterogeneous data, need to be addressed
- Planned to improve the ability to handle large in-memory extracts by replacing TDE data file formats with an in-memory engine named as Hyper
- Provision of Event-based scheduling and conditional alerts to be featured
- Development of API interface services for better extensibility
- Reduced complexity in packaging to perform large deployments of lightweight customers
Recommended Articles
This has been a guide to What is Tableau. Here we discussed the Key Concepts, Working, Feature, and Products of Tableau. You can also go through our other suggested articles to learn more –
