ggplot

Was watching a YouTube video on time series in ggplot2, moved on to another YouTube video on ggplot which turned out to be in Brazilian Portuguese.  In which, it turns out -- I could probably have worked this out if I had thought about it - g is pronounced like the j of French je, and the final t is pronounced (roughly) tchi (because no word can end in a consonant in Br. Portuguese) - in other words, jéjéplotchi.  Impossible not to love. When the dog barks, when the bee stings, when I'm feeling sad, I shall think of jéjéplotchi and not feel so bad.

Sapir-Whorf and ggplot2

To some degree, we are constrained in our ability to solve problems if we only know a single language. This situation has been recognized different ways by the programming community. The Logo programming language was built based upon constructionist learning theory and was intended to provide a “mental model” for children to come to understand mathematical constructs. In recent times, many programmers have committed to being polyglots, learning new languages as a part of professional development. Their concern is not always to learn the latest language that they will need to work, but to find out new ways of conceptualizing problems and structuring solutions.

R Chart on language and thought and ggplot2

head to head

Over on Learning R, the intrepid RLearner is going through Deepayan Sarkar's book on data visualization using Lattice and replicating the graphics using Hadley Wickham's ggplot2. It's completely enchanting.

Best excuse ever

My dear friend Rafe Donahue has sent many helpful suggestions with regard to ggplot2. He comments:

Don't think of the ± of 0.2 to each of the data at the repeat
locations as modifying the data.  You are not modifying the data, you
are providing a different plotting algorithm for those points that share
a location with other points.

a line of argument which it is hard not to love. I'm not saying it's not valid, no... and yet I see myself, down the years, explaining innocently that I was not actually modifying the data as such, I was just providing a different plotting algorithm etc. etc.

Meanwhile Hadley Wickham very kindly sent the correct line of code to change the y axis. I type this in, and by the simple procedure of providing a different plotting algorithm for those points that share a location with other points produce

which really is terribly nice. Further information on ggplot2 is available here.

Readers who have not spent much time with Excel charts may be inclined to accept uncritically the complaints of PP; a wealth of information on what can be achieved is available at Peltier Technical Services, here.

On the subject of providing a different plotting algorithm for data whose points overlap, Rafe reminds me that

As I said before, we did this in the
baseball plot data y putting them in little boxes.  Of course, you need
to decide on the size of the box, etc, but that is the price to pay.

For those who missed the great bivariate baseball score plot the first time round, this enables the user to select a team and one or more aspects of its game history and generate a bivariate plot, for example



You can create your own plots here.