Now we've looked at categorical and numerical data types. But I told you there was a second way to classify data types, and let us to see them as either being discrete or continuous. Let's start of with discrete data types. Now a discrete data type contains values that have a definite, minimum size. They cannot be further subdivided. A classical example is just the roll of a die. Now, if it's a six sided die, it can then either with one, two, three, four, five or six face up. It cannot land with one and a half, so that is a discrete value, those are discrete values. Discrete of all the possibilities are finite I can count them, one, two, three, four, five, six. Now I want to caution you against two variables: one is money and the other one is time. Now, in most currencies, we have a minimum currency value, one cent, one penny, whatever it is. We can't have a quarter, half, eighth, or 16th. That's not really true, though. If you think about large institutions working with large amounts of money. If you took some percentage value, you could run into many decimal vales, and you can't throw all of those away, because you'll be throwing a lot of money away. The same goes for time, we could say that one second is the minimum value, but you can certainly sub divide, sub divide, make smaller, make smaller. So specifically money. Even though it feels intuitive to call money discrete, or continuous data type, it is a continuous data type. The set of possible values are really infinite, I can always subdivide, then make them smaller. Divide them in two, divide them in two, go deeper, get more and more decimal values. Now it's not always so practical though, so we don't really want to be too pedantic as to what really defines a continuous data type, we want to be slightly practical about it. Let's look at systolic blood pressure. If we had a blood pressure cuff And, your data says 120 millimeters of mercury. But I can have a better cuff that can tell me it's 120.5. What, theoretically I could have even better piece of equipment and it could tell me 120.555673254. That really just depends on how good my piece of apparatus is. So, systolic blood pressure type would be a continuous data type. Although for practical use we're always ever going to use integer values. If we look at white cell count, we might have a white cell count of 9.9 times 10 to the power 9 per liter. Now you might tell me, well there is an absolute minimum. I can't count them because I can only go down to one cell. I can't have half a cell. Under these circumstances, we did it with such large numbers, something to the power nine, such a large number. But, for all practical purposes, and for mathematical purposes, this is a continuous data type. So there you have it, all sorts of ways to classify and categorize our data points. And the reason why we did is that we have statistical analysis for all of these different types.