About me Blog

Interactive data exploration using Dc.js

⏳ 1 mins read time


Interactive visualizations are a great alternative to static figures for data analysis. The reader is able to define his own story by making its own explorations and get its own conclusions. In this post I show a simple way to get interactive visualization using javascript functions dc.js.

The code can be found at the following gist and the visualization is available at this link. The visualization have to main files:

  • index.html that contains the front end of the visualization
  • graph.js that contains all javascript that runs the interactive components

Front

I used bulma.io css framework to set the differents charts in a responsive way. This framework provides a tiles disposition to create a grid of content.

The structure of the grid in terms of bulma instructions is the following:

tile is-ancestor
|
β”œβ”€β”€β”€tile is-vertical is-8
|   |
|   β”œβ”€β”€β”€tile
|   |   |
|   |   β”œβ”€β”€β”€tile is-parent is-vertical
|   |   |   β”œβ”€β”€β”€tile is-child
|   |   |   └───tile is-child
|   |   |
|   |   └───tile is-parent
|   |       └───tile is-child
|   |
|   └───tile is-parent
|       └───tile is-child
|
└────tile
    |
    β”œβ”€β”€β”€tile is-parent
    |   └───tile is-child
    |  
    └───tile is-parent
        └───tile is-child

Back

The div elements that are inside the tiles are created using the graph.js file. This file uses d3.queue to load the json file ./df.json using d3.json function and wait until it finish to execute the function makeGraphs.

The function makeGraphs:

  • Parses date time data
  • Creates a crossfilter and add the records. A filter represents a multidimensional dataset.
  • Definition of the dimension of the crossfilter.
  • Groups data on each dimension.
  • Define the charts and the id to be used in the html file.
  • Update the heatmap if any dc chart get filtered

The barplots are made using standard dc.js api. They are connected through crossfilter.js so when each dimension each filter, the results can be see in the other dimensions. In addition, the Leaflet map is connected to the charts so when one chart is filtered, the results are shown in the heatmap.

Conclusion

The final interactive visualization is available in this block link

You may also like: