The next method we need to introduce is change mapping. What is change mapping? The basic idea is that interactions that change the mapping are interactions that affect the way data is translated to a visualization. To a visual format. So the encoding rules that translate data attributes and data items into visual marks and visual channels. So what is interesting about this type of interaction, is that it can lead to changes in the parameters of an existing graph, or more dramatic changes that basically lead to a reconfiguration of the visualization, so that a completely type of graph is used. In order to make this clearer, let me give you another demo that is based on the same scatter plot example that I've shown in some of the previous videos. So here, what we have is the food data set that I've shown before. In this kind of plot, every single dot represents one food product. And each food product here is mapped according to three main attributes, the amount of carbohydrates, the amount of calories and the food type. And carbohydrates is mapped to the x position of the dots, calories is mapped to the y position of the dots. And food group, or food type, is mapped or encoded with color hue. Now, let's see what kind of interactions I can use, that actually change some of the mapping rules. Let's start with something really simple. Let's say that rather than using a dot like the one that I'm using here, I may want to use a different type of mark. So, for instance, I could use a square or a circle. So that's an example of a slight change in the encoding rules. Where the original encoding rule is encode every single food product with a dot and I change it to square or circle. So, let's try something a little bit more complex. Let's go back to circle. Let's say that I'm no longer interested in showing the relationship with food group, but I'm more interested in showing the relationship with the amount of water that is contained in a food. So I can remove the food description attribute from color and use water. And say I now want to encode with color, the amount of water that is contained in the foods. And as you can see here, interestingly, not only we changed which attribute is encoded with color, but also what color map is used to represent this type of information. In this case, since we switched from a categorical attribute to a quantitative attribute, we need to use a different color map. Now, let's remove water and let's say that I want to go back to investigating the distribution of different food types, but rather than showing this information directly or only with color, now I want to split it into a small multiple visualization, where every single scatter plot represents one food product, one product type, sorry. So how do I do that? Well, I can take food group, drag it here and now I'm generating a separate scatter plot for each specific group. Let me change the visualization here to make it clear. So this is a little clear. Let's make the size of the dots a little smaller because we have less space here. Let's go back to this. So that's a different configuration. We're still using scatter plots, so it's somewhat similar, but in a small multiples fashion. And as you can see, there are different patterns that emerge from this visualization. And this is another example of changing the mapping rule. I'm no longer mapping data, I'm no longer positioning the dots just in one scatter plot, but in multiple scatter plots, according to a specific rule. So this gives you a sense of the fact that changing the mapping can actually lead to dramatic changes, but these changes can be very useful for investigating different types of questions in a dataset.
