MapInfo Pro

There are a number of user scenarios where raster grids can be used for more than visualisation of elevation data. For example, creation of a population density grid can be used as a visualisation but can also return a population metric for each cell. This data can then be queried to hotspot area above or below a particular population density. 

In another example, a grid could be used to improve my understanding of spatial distribution of retail outlets across a country, this would be useful for competitive analysis.

Let’s take a UK points of interest dataset. It represents the majority of known chain pubs across the United Kingdom. As a simple point dataset, it gives a good initial indication of the spread and density of pubs across the country.  See Attachment

However in the urban and city areas it becomes difficult to identify just how densely clustered these pubs are. It becomes a case of not being able to see the woods for the trees.

An improved and more accurate display of this data can be created using an interpolation as outlined below.

Raster > Create Raster > Hotspot Density. 

Enter the settings, many of the default and Automatic options are very good for a first pass at the interpolation.

See Attachment

Once the density map has been processed it will be displayed in the map. This gives a hotspot view of the density of the pubs. In order to focus the map on the areas that truly have a high density, colour and colour stretch changes are required.

Raster > Colour > Choose a colour ramp that colour the less dense values as white.

Raster > Colour > Colour Stretch > Linear 5-95%. 

Histogram and Linear Stretch along with their clip options are designed to stretch a smaller range of values (the raster grid values) over a wider range in order to accentuate contrast between features of interest.

The result of the type of change can be seen below. The image on the left show’s a less refined pattern. In appears that much of the UK has a high density of pubs. In reality after some manipulation of the colour and histogram, a much more realistic story appears. In regard to making informed decision making based on location, the second map gives a much more accurate and precise set of areas.

See Attachment

Love this Chris including the subject matter! Can you give a few more examples of when creating a raster allows for better understanding? Not just the use cases but the analytics tools/technique that can be applied to a raster?

Also I'm looking forward to your webinar on this subject May 24! The following link is no longer available. To access this information, copy the title found in the url into the search tool on the top nav bar, or try searching for key words in the post. https://li360.pitneybowes.com/s/detail/a7x800000008R0FAAU

Great use case Chris! Also I really like how you elaborate on the display settings to convey the result of the analysis in best possible way.

Another way to make a raster from a point file is the Distance method under Create Raster. Here is an example where service points (could well be pubs) are being used as input and the resulting raster reveals which areas have high availability of services and which ones don't. Again, the raster aggregates information from all the points in the layer, and creates a visualization that is more seamless than what you would get in a simple buffer map.

See Attachment

This is more convincing.