Skip to content

Visualizations

May 10, 2013

thousandwordsThey say a picture is worth a thousand words. An image can convey an incredible amount of information, though the interpretation can vary widely depending on the viewer’s perspective. However images can often condense information into a much smaller space. One that is often clearer to many people than a large spreadsheet or numbers or pages of words.

Visualizations can help us understand and consume a large amount of data quickly. They don’t always provide all the information we need for a decision, but they make it easier for us to understand where we should look deeper. There’s a great piece on how visualization tools help us understand big data, and also a tour of visualization methods (thanks to Brent Ozar, PLF).

Many of us are working with larger and larger amounts of data, and being asked to help business people find tools or methods to work with data. We’re also writing lots of reports to extract information, and I wanted to ask this question:

What visualization methods do you prefer working with?

If you have more than one, let us know, and if you have preferences for certain types of data or analysis, let us know. If you’ve found some techniques that don’t work well or mislead people, let us know about those as well.

Images, animation, color, and more can really improve the transfer of information, but it’s not always as useful as we might think. One of the points brought up in the editorial linked above is that our tools and visualizations need to account for uncertainty in the data quality. I don’t know what methods and tools we’ll use in the future, but I do know that our choices continue to grow in the SQL Server space as new tools like Power View and Geoflow are released.

Steve Jones


The Voice of the DBA Podcasts

We publish three versions of the podcast each day for you to enjoy.

About these ads

From → Editorial

Comments are closed.

Follow

Get every new post delivered to your Inbox.

Join 4,299 other followers

%d bloggers like this: