Seeing how things change over time can be important for a business so that you can figure out what works best. Square, the company that turns your iPhone into a credit card reader, just released Cube, an open-source system to help you visualize time series data. It’s built on MongoDB, Node, and D3.
Once you’re all setup server-side, it’s pretty quick to put together a dashboard, as shown in the video below:
A lot of the boiler plate time series work is done for you via the visualization components, but if you want to dig deeper, you can also access Cube’s query and aggregation functions. This gets you sums, medians, minimums, maximums, and some others, as well as clumping by minutes, hours, days, weeks, or months. If you’ve done this sort of a stuff with your own code before, you know how annoying it can be.
Cube is still a work-in-progress, but with Mike Bostock, now a visualization scientist at Square, this could be something to keep an eye on.
[Cube]
but….. why? did we have a problem of not being able to (easily) represent data over time?
The representation is easy, but the stuff in between can be a hassle sometimes, especially for beginning programmers.
I think the context of why this was created was explained in Jack Dorsey’s keynote at Stanford earlier this year. See http://ecorner.stanford.edu/authorMaterialInfo.html?mid=2635 if you haven’t already watched it.
As you probably already know, Jack is the founder of twitter. In his speech, he talks about things they put in place at twitter so they could monitor whats going on (like server capacity and operational status). Jack mentions that they are doing the same thing at Square. But, he’s giving this monitoring software to his customers that use square so that they can visually see the money they’re making through his authorization service. I believe his focus is mom-and-pop type stores, and this is something that they’ve never been able to visualize before (other than looking at the good ‘ol accounting T-chart…) Well done, Jack!