Objective : To obtain a global view on the combination of several indicators and entities. Ideal for cross-cutting data.
A "Heat Map" is a data visualization technique. It is very useful for the visualization of complex statistical data.
Specifically, it is a data analysis tool based on the use of colors. They substitute the height or length bars on a diagram.
An example of ToucanToco's interactive Heat Map. Identifies vacancies and assesses needs - Gives the department the opportunity to align staff skills with initiatives and objectives
Heat maps are used in all areas to make data sets understandable and usable. For instance, Heat Maps are used by doctors, engineers, sociologists, or researchers in all fields.
They are also used by webmasters to visualize which parts of a website attract the visitors' attention the most. The areas most viewed are represented by a warm color, and the least viewed by a cold color.
In general, heatmaps allow you to highlight the relationship between two variables. This makes it possible to identify trends and patterns intuitively based on colors.
This can be any type of variable, such as category labels or numerical values. Colorization on the other hand can represent any kind of metric. It can be a frequency count, or a statistical summary such as an average. It can also represent a quality instead of a quantity.
When creating a heat map, choose the color palette that best matches your data. As a general rule, bright colors represent low values and high colors represent high values.
Don't forget to add a legend to your heat map to specify what colors correspond to what. You can also add the corresponding numerical values for a more detailed result.
There are several types of heat maps. One of the most popular is the clustered heatmap. It allows you to create associations between data points and their characteristics. Analytical tools offering this type of heatmap implement clustering in their process. This type of map is very useful in biology, especially to study the similarities between several individuals' genes.
Another variation of the heatmap is the "correlogram". Each variable on the two axes is replaced by a list of numerical variables in the dataset, and each cell represents the relationship between the intersecting variables. This type of heatmap allows analysts to understand the relationships between variables to create predictive or descriptive statistical models.
Another example of interactive HeatMap available in our data visualization tool ToucanToco