hw3 Review

[CanyonFoot]

@simonpcouch

[simonpcouch]

5.1: The challenge here was to differentiate between the number of home runs hit and allowed in the graph; not to graph a ratio. That step was the one that required gather(). This way, both hit and allowed could be graphed separately, differentiated by color. Otherwise, clean and efficient code, though.

5.2: Yup, this looks great! Efficient code, and really helpful that you put an example in to test it.

5.7: Super well done on this one. For one, recreating the dataset made for testing the code really easy.

5.15: I also switched the "F" values to female, good looks on that one. Really helpful in terms of computing power to select only the years and names that had both a male and female entry. The way you did it was really artful/creative, too.

Generally considered bad practice to pipe into an entry on the same line. Also, saving the intermediate steps (i.e. babynames1) might not be a good idea when working with datasets as big as this one.

Really cool idea on how to filter the rare names! Last few lines are efficient and intuitive.

5.16: Any reason on saving the intermediate steps? As far as I can tell, breaking the pipe doesn't seem necessary.

Also, gather() and spread() are intended to do what you did with dplyr (ending in the full_join) but in many fewer steps. Your plots get back to the same thing I mentioned with 5.1; the goal of this is to tidy the data so that both geom_lines can be on the same plot, differentiated by color (or some other aesthetic). A final gather() right before piping into the plot would make the difference here. As well done as is possible using dplyr though!

[CanyonFoot]

Re: 5.16, thought I about doing it the way you described, but I understood each month to represent an observation, and therefore felt that it was appropriate to make the columns the surface and bottom temperatures. I split up the dataset and joined it just to make it a little easier to do this. Additionally, the choice to create two plots was intentional, as you can see, the plots are similar, and even using color and alpha I found it very difficult to interpret with both lines overlapping so much.