• 131
• 255
• 14

I got to do some SandDance visualization

by bernt & torsten

Scrolling my news feed this morning, there was one thing that caught my eye, the news I read was that Microsoft had open-source their visualization data explorer tool SandDance.

I got out of bed and headed over to my home office, powered on my laptop and downloaded the SandDance tool from Github. I cloned the repository source code to my laptop. The interesting parts of SandDance visualization data explorer tool are that you can use it to find insights in your data and tell data powered stories supported by your data, you can use it to build use cases based on evidence in your data set, test hypotheses, dig deeper into surface explanations.

How to use SandDance

SandDance consists of several components programmed with JavaScript and the React framework:

• The SandDance core visualization canvas.
• The SandDance-react core SandDance visualization canvas for use in React-based applications.
• The SandDance-explorer core SandDance visualization canvas with UI to enable data exploration

You can embed the components into your current App infrastructure or you could just get started to use the out of the box explorer, you need node installed after that you just need to install node for the SandDance project and start the SandDance explorer.

Exploring data

You can run the SandDance explorer in light mode or dark mode, I prefer the dark mode, you need a decent dataset for that to show well. The explorer comes with two sample demo datasets, you can, of course, use your own dataset. The supported data formats are json (the default), csv (comma-separated values), tsv (tab-separated values), and topojson. When you use your own data your file will not be uploaded, it is only used by the browser on your computer. Below is an example of the Demo Vote data as a scatter chart.

With my own dataset

I had a small dataset with data collected from a sea trip, it had various data points like temp, wind direction, weather condition, latitude, longitude, barometer pressure etc., I went ahead and loaded the data and started selecting different charts to see what I could get out of the data. I created a column chart, with what wind-force with barometer data, and I got some interesting insights, as shown in the chart below.

Due to my small dataset, the scatter chart below shows the wind force at a specific latitude and longitude during the journey.

Changing the chart options are very easy, in the below Bar chart I’m showing the number of data points at a certain latitude and what the wind force was at these latitude data points. It was very easy to switch options and also change the chart type.

I just started to scratch the surface of the tools, what I want to do next is to incorporate it in one of the node.js data projects that I’m working on. That will be an interesting development project for me.