It is possible with minimal effort in InstantAtlas to throw a set of statistical indicators into an interactive dashboard and view them using the default maps, tables and charts. While that approach is fine in many cases, my interest lies more in how you can delve into the toolbox of design options to create visualisations that are highly tailored to the data they show.
There are many chart types available in InstantAtlas to help you with your data visualisation, from bar charts and line charts to radar charts and bubble plots. In this post I’d like to feature the Area Breakdown chart. This is a bar chart that shows a breakdown of indicators (or associates) for the area (or areas) that the user selects in the map or table. It has a unique characteristic that sets it apart from the other bar charts in InstantAtlas: it is possible to assign a different colour to each bar. You could therefore shade each bar according to political party for example, if the bar chart is showing election results.
Colour can be used in all sorts of ways to help data interpretation. I liked the idea of using the colour of the bars in the chart to represent the time of day, one possible scenario being display of how different crime types vary over a 24 hour period. Visualisation of the timing of crime has been done before for US cities, and even in some detail for Chicago. However the City of Chicago data portal provides access to such a fabulous crime dataset that it is too good to pass up. Downloading the over 6 million reported incidents of crime going back to 2001 did seem overkill for this experiment so I decided to limit my dashboard to incidents reported in 2015.
Rather than burden the dashboard with numbers I kept this to just three key stats to leave plenty of space for the area breakdown chart. For the chart itself I used a black-to-yellow gradient to simulate the change in light intensity throughout the day.
To cover myself I will say that this example does contain some somewhat gratuitous use of graphical elements that are superfluous for interpretation of the raw data. But every data viz geek likes to break the rules from time to time!
Click HERE to view the report.
In 2015 ONS released an interactive Atlas of Tourism for England and Wales created using InstantAtlas. This was followed by a statistical bulletin that classified the areas in the atlas to show the importance of tourism.
The template that ONS chose for the Atlas of Tourism is the Double Map Time Series, which allows the user to compare the data distribution for two variables using a map and bar chart. In order to demonstrate the versatility of InstantAtlas I have created an alternative representation of the same data using our Area Profile template. It opens showing the spatial classification developed by ONS, with clusters 4 an 5 representing what they term ‘holiday hotspots’. The report allows the user to select specific authorities in the map in order to profile them using the ‘spine chart’ graphic on the right – and to click on the indicators in the spine chart to update the map. For benchmarking purposes, the median authority has been included in the spine chart. This report is more geared towards a user that is interested in a particular authority and its tourism profile.
Click on the screenshot below to access the Tourism Area Profile report.
Pierre Jenkins, Head of InstantAtlas Support
The data in this report were downloaded here and are licensed under the Open Government Licence v3.0.
In Dec 2015 the BBC reported on Britain’s busiest and quietest stations. They made use of data released by the Office of Rail and Road (ORR) that records the estimated number of entries and exits based on the total number of people traveling from or to the station. The list of busiest stations in the BBC article isn’t particularly surprising – London’s main stations are by far the busiest (Waterloo, Victoria, Liverpool Street, etc).
What is more interesting to me is the year-on-year change for each station – whether usage significantly increased or decreased – and seeing this information on a map where it belongs. The ORR publish a time series dataset for 1997-2015 that includes Ordnance Survey grid references. This makes it an easy job to plot the location of the 2,608 stations in InstantAtlas.
My report shows year-on-year change. One challenge was the large number of points (stations) that obscured one another and made the map hard to read. To overcome this I filtered out stations where year-on-year change was less than 50% and experimented with the size, colour and opacity of the points to strike the best balance.
View report: Dark or Light canvas.
Click the Play button to animate year-on-year change, or click the dates at the foot of the report to jump to a specific year.
Note that the 2003-04 date carries a * flag. The ORR did not provide estimates for 2003-04 and so estimates for this year were calculated as the average of the years to either side.
The Profiles module of InstantAtlas Server provides a powerful set of tools to design and deliver highly effective reports containing a mix of appropriate information relating to a specific area or number of areas. They support a wide range of chart types and tables, intelligent text, and images. This tutorial is aimed at InstantAtlas Server (IAS) administrators that wish to enhance their profile reports by adding an interactive location map.
Including a map at the top of a profile that shows the location of the area being reported on is useful for obvious reasons. An end user that is not familiar with the area will want to know where it is. A map also makes the profile more visually interesting, particularly if the rest of the profile is dominated by text and/or tables.
InstantAtlas is great for showing statistics by area and there are many examples of reports containing area-based maps (called choropleth maps). But IA is also suitable for creating point maps.
One excellent example is this report published by the Northern Ireland Statistics & Research Agency to show traffic collisions in Northern Ireland in 2012. I’m a big fan of this report and will use it to illustrate my top 5 tips for creating point maps with IA.
Traffic collisions report for Northern Ireland 2012
InstantAtlas Desktop version 6.7.1 – released earlier this month – includes a much-improved Feature Card component. Our development team have been slaving away on this mysterious component that wields great powers.
In interactive mapping visualisations such as InstantAtlas the humble legend gets a makeover. But just as with more traditional static maps, the importance of getting the legend right is crucial. Here are my 5 top tips for creating effective legends in your InstantAtlas reports.
Read the full article