So in this week we talk about visualizing network data. What is network data? Well, we will see that network data is about datasets that describe connections between objects or relationships between objects. But before I talk about that, I want to briefly talk about the type of data that we have seen so far, so that I can highlight what is the difference between the datasets, the type of data that we have seen so far and the type of data that we're going to analyze in this week. Okay. So the most classic type of data that we have seen so far is tabular data. Always tabular data- what is the shape of tabular data? Where typically you have a number of objects and for these objects you have a number of variables or attributes that describe the properties of these objects. These values typically are either some categories or some quantities. And this is what we refer to as Tabular Data. A little variant over tabular data is temporal data. So typically, it's very similar to tabular data with the addition that one or more of the attributes that you have in the table is time or date. Then, we have seen geographical data. And what is geographical data? Again, basically, very similar to tabular data with the additional information that the objects may be geographical objects and categories or quantities that are contained in the data are also geographical categories or geographical quantities. Okay. So, now we switch our attention to network data. It's not little switch, because there some fundamental changes between the type of data that we have seen so far and network data. What is the main difference? Well, the difference is that, not only we have objects and values but most importantly we have relationships within these objects. So, let me try to make this more explicit. In a network data we have objects that are expressed, that are called Nodes and then we have links also called Edges that connect these nodes and represent some sort of relationship between these nodes. The way these type of data typically looks like is that we have one table that represents the nodes, where every single row is one node. Typically, you may have one or more values associated to this nodes. You can also have zero values just the list of nodes in some cases. Then, you have another table or list that represents the links so, every single row represents one link between two nodes. So as you can see in this table here, we have a link between A and B, A and D, A and C, A and E and so on and in some cases we can have one or more associated values to each link. So, that's the main difference. The main difference is that with network data we have basically two sets: Sets of nodes and a set of links that connect these nodes. So here is an example of potential real-world data set of friendship network. So I think everyone understands today what a friendship network is, you can imagine your network from friends in Facebook or in real life. Right? So, every single node is a person and the connection, a connection exists between two persons if they are friends. Okay. So, in this case in this image we have John, Jessica, Paul and Mandy. John, Jessica and Paul are all connected but Mandy is only connected with Paul. I also put some numbers in the nodes, that in this case represents let's say age just to give you a sense of the fact that some values can be associated to the nodes and in principle you can also have values associated to the links or edges. I want to conclude by saying that, there are lots of real-world phenomena that are described or can be described through networks, many of them. So, as I said association between peoples in friendship network for instance, some people study, connection between animals say animal migrations between one region and another region of the world so you have a network there. In biology networks are very common any kind of biological system that shares some properties let's say DNA or elements of the DNA, connection between companies, between Cities, transportation networks, there are many real-world phenomena that are described by networks, and that's the reason why they're so important and the reason why we need to figure out effective ways to visualize them.
