That is, in 1985 there was a greater percentage of countries with a summary index less than 6 than the same percentage in 1970. The second observation is that in 1985 even though the number of countries with some economic freedom increased (more data points), the variability in summary index values is smaller due to how its ECDF curves. One is that 1970 appears to have the least amount of data points, which may indicate that very few countries had economic freedom or some quantities weren’t recorded. There are a couple of other interesting observations with 19. Therefore economic freedom increased overall from 2000 and at least in 2015 also. The immediate observation you could make is that the summary indexes were overall lower in 19 versus the later years. What observations can you make from this plot? Try to think about this before reading about them in the next paragraph. Much clearer isn’t it? Plus this allows you to see distinctions in the variation of the summary indexes over 15-year jumps. Let’s take a look at the ECDF chart above in the post.ĮCDF plot of summary indexes for 1970, 1985, 2000, and 2015 The ECDF essentially allows you to plot a feature of your data in order from least to greatest and see the whole feature as if is distributed across the data set. So what is an Empirical Cumul….What’s an ECDF?Īn ECDF is an estimator of the Cumulative Distribution Function. The code isn’t included in this post so you can focus on the visuals and interpret them.

The code used for these charts and a copy of the excel file are in this Github Repo. That data was also used to make the plots above.

In this post, we will explore what an ECDF is, why to use it and the insights we can read from it using our Economic Freedom of the World dataset provided by the folks at #MakeoverMonday. It is known as the Empirical Cumulative Distribution Function (try saying that 10 times fast…we will call it ECDF for short). Scatter plot comparing the Economic Summary Index of Countries with their Size of Government.īut there is one summary statistic visualization that I did not learn about until I explored a statistical thinking course from Datacamp.

Why do teachers use cdf files