How FusionTime Works

FusionTime is a JavaScript library that helps you visualize and explore time-series data interactively. It supports data with atomicity ranging from milliseconds to years. Along with the capability to smartly aggregate data on the time axis , it also has built-in powerful components (like time navigator , range selectors , event markers , etc.) which enables easy and intuitive exploration of large data over a long time period and smart annotations.

In this article, we will discuss the complete flow starting from connecting data from the data source to rendering the chart.

How FusionTime is different from FusionCharts

While FusionCharts can help you visualize data in dozens of different ways, FusionTime works exclusively with time series data. Unlike other varieties of charts in FusionCharts, which only (till now) accept data in JSON/XML, FusionTime accepts data in rows and columns as a Datatable . You can then apply different data operations ( filter , grouping , etc.) on the DataTable to optimize your data as per the requirement.

FusionTime Data Layer

So, how do you create the DataTable? The FusionTime data layer comes with a DataStore which accepts data in JSON/2D array format. The data fed in the DataStore is converted to a root DataTable using the createDataTable() method. This DataTable contains the source data in a tabular format.

All you need to do is provide the source data in JSON/2D array to the DataStore and the data layer takes care of the rest. You can then either provide this DataTable to render the chart or apply data operations to create new DataTable with optimized data.

The DataStore also contains a set of data operators (filter, grouping, etc.) which can be applied to any DataTable. Any operations applied on a DataTable to optimize data generates a new child DataTable. Likewise, any number of data tables can be generated from either the master or other child data tables. You can then provide any one of them to render the chart.

To help understand better refer to the illustration below:

FT Works

The above illustration is a flow diagram which defines the following:

  • Source data transfered to the browser in JSON/2D array format from the data source.

  • Create a DataStore and load your data to create the root DataTable.

  • The data is then fed in the DataStore. To identify the data you need schema and data .

  • The root DataTable is then fed to the chart renderer along with chart configurations to render the chart.

  • Alternatively, you can also apply data operations on the root DataTable to generate another child DataTable.

  • You can then feed the derived DataTable to the chart renderer along with chart configurations to render the chart.

Now that you have the basic understanding of the data flow for FusionTime it's time to install and build your first chart .

Was this article helpful to you ?