Tableau Public: Gangatharaprabu G - 10BM60026 Deepthi B.V - 10BM60095
Tableau Public: Gangatharaprabu G - 10BM60026 Deepthi B.V - 10BM60095
Introduction
Tableau Public is a cloud based visualization tool which is currently majorly used by bloggers, journalists etc. It is the latest offering from Tableau SoftwareTM, which is a cloud based Business Intelligence platform. It allows users to share data visualizations interactive charts, graphs, maps and the like, on the web. It also allows results to be shared by social networking and other websites
Some users
Features
It has access to files in Excel/Access format as well as different text file formats. Tableau Desktop Public Edition is a Windows software only. The data will be stored in TableauServers and not on personal computers, and any changes done will have to be done online. It can also publish data to your iPad in addition to the web, which makes data analysis more social and a mobile experience.
Advantages
It is possible to link different views of the data and let the user select the subset of data he is interested in. Unlike few other visualization tools like Protovis, it doesnt need the user to know programming. It offers powerful tools in filtering and selecting data which offer flexibility to the user in creating effective views of the data. The Show me tool box- When the dimensions and measures of interest are selected ,Tableau suggests visualizations that best suit the data.
Disadvantages
It has to store data in a third party host, TableauServer. Therefore during high traffic, it gets slow (Since everything is done in a browser). It can work up to only 50MB of data. It cannot show completely custom views.
Data Sources
The following two datasets have been used in this tutorial.Sources of the same have been mentioned in references. RetailData ForeignAidData In addition to the above, interested people can go into the following sources to find more datasets, for analysis.
Connecting to Data
Connect to data opens the dataset for analysis. Data can be of any of the following formats.
Formats allowed
Formatting data
Before opening any dataset for analysis or for creating dashboards or posting them to blogs, it has to be formatted to be conducive to analysis and compatible to the software format. Here are a few basic things one might look into The dates should be in the date format(dd-mm-yyyy) The first row of the should be column headings
Creating visualizations
Bar charts Maps
Demo
Data used: This data consists of government aid given to different sectors across different years in different countries.
Source: http://www.tableausoftware.com/public/training
Creating Map view If we double click on a geographic dimension(Ex. Country name), a map is generated with the system generated latitude and longitude values. Color and Size fields Unknown locations
Map options
Other visualizations
Data used Retail sales across years
Tables
Bar charts Pie charts Histograms
Analysis
Creating Dashboards
Go to Dashboard New dashboard. Add the sheets to the left dashboard pane. Layout of the dashboard can be modified by the user
Using filters
Filters can be used for selective visualization Filters can be used in Views Dashboards
For the previously discussed Foreign aid example, if filters are added and country name is selected to be the filter, the visualization is as shown.
The bottom right pane shows the list of countries.
Types of filters
Multiple list : this is the default filter. As shown in the above example it shows the names of all the countries. Single valued list: The list just shows single country. The output is shown country by country. Custom list: It shows the list of countries as a search box Compact list: It shows the list of countries as a drop down list
Slider bar: The slider bar allows the user to change country.
Filters in dashboards
Business Value
Tableau as we have seen is very much useful in visualizing data both local and web based. It is possible to create dashboards and track KPIs For example : If we want to track the effectiveness of a ad campaign on social media It is possible to track all these data Location wise viewers Total number of visits Pages along with which it is visited Increase/decrease in number of viewers over a period of time It is also possible to construct a dashboard of these information together We can also filter and deeply analyse the particular area, particular time of visits etc Also if needed and if it exists, trends can be constructed using regression There can be many applications of tableau in various domains like the above example