Web > Panels

Displaying Data

There are three types of panel in AMI:

1. (Static) Visualizations
2. Realtime Visualizations
3. HTML Panels

Adding a new panel generates a blank panel that can be populated with these three options. The first two options are 3forge's primary methods for turning data into visualizations.

HTML Panels are generally more customizable and have a wide variety of interactive fields that users can configure. These not only change the look and feel of dashboards, but can contain analytical functionality too. See the Styling pages for more information on HTML Panels.

Datamodel or Realtime Visualization?

AMI uses both static and real-time visualizations for displaying data in AMI Web. How and when to use them can be broadly summarized as follows:

1. (Static) Visualization

   Static tables and visualizations are backed by datamodels. Datamodels are tools hosted within AMI Web allowing users to query and manipulate data from several data sources. See this page for a step-by-step guide on using datamodels.

2. Realtime Visualization

   Realtime visualizations are optimized for real-time visualization. They take data directly from tables in the Center and display them, without additional computation or data manipulation.

   Note

   While realtime visualizations are optimized for real-time data, datamodels can also be used for realtime visualizations.

   See Realtime Datamodels for more information on handling real-time data in a datamodel with conflation.

Generally realtime tables are faster, but datamodels provide greater flexibility and the ability for complex queries and data manipulation. See this table for a full comparison.

Note

Both static and realtime tables provide fully customizable columns. Users can freely edit and add columns, even after table creation, without affecting the underlying data.

Static vs Realtime Tables

The table below presents a deep dive into the tradeoffs between the two:

Topic	Datamodel	Realtime
Flexibility	Datamodels can use arbitrary AMIScript to manipulate data in any way.	Realtime visualizations can only use data from the Center subject to simple column-based manipulations (e.g. addition, ternary operators) but not more complex manipulations (e.g. joins).
Datasources	Datamodels can use data from any source.	Realtime visualizations can only use data loaded in the Center.
Scalability by no. users	Datamodels run on a per-user basis, meaning if there are 10 users logged in, then the datamodel must run 10 times and keep 10 copies of the data.	Only one copy of a realtime table is kept in the Web, meaning it can scale to an arbitrary number of users.
Scalability by data quantity	Datamodels keep a per-user copy of all of tables produced by the datamodel.	One copy of a realtime table is kept in the Web.
Run	Datamodels are run whenever they're explicitly told to run.	Realtime tables run whenever a row in the underlying table is inserted, updated, or deleted.
Queries VS Delta	Datamodels run on queries. Every time a datamodel is run, it must do all of the work described in the datamodel. This means if we pull 1,000,000 records, then re-run the datamodel and pull those same records plus 1 more, we will have done duplicate work on the 1,000,000 records.	Realtime tables run on deltas, meaning an incremental change to the data will only require an incremental piece of work. If we have a realtime table with 1,000,000 records and add another, then we'll only need to do work on the 1 new record.