
Introduction
Various statistical distributions have been invented, which are the equivalent of the wheel for data analysts. Just as whatever I think of comes out differently in print, data in our world doesn't follow strict mathematical laws. Nevertheless, after visualizing our data, we can see that the data follows (to certain extent) a distribution. Even without visualization, we can find a candidate distribution using rules of thumb. The next step is to try to fit the data to a known distribution. If the data is very complex, possibly due to a high number of variables, it is useful to estimate its kernel density (also useful with one variable). In all scenarios, it is good to estimate the confidence intervals or p-values of our results. When we have at least two variables, it is sometimes appropriate to have a look at the correlation between variables. In this chapter, we will apply three types of correlation.