Problem 1

  1. load this dataset on scapula morphology in Pan, Homo and Gorilla, and the Dikika fossil child. Details on this data can be found in Alemseged et al, 2006

  2. Calculate the natural log of all the numeric variables. Save these in your existing dataframe.

  3. Make a subset of the data including only the extant species (G, H, or P, representing Gorilla, Homo and Pan), excluding the Dikika fossil (D).

  4. Perform a discriminant function analysis (DFA) on the data. Which variable has the strongest loading on LD1?

  5. Make a beautiful plot (using ggplot) of LD1 versus LD2. Hint: to get the LD scores, use the predict() function on your discriminant function analysis model object, as shown in class.

  6. Plot the Dikika fossil on the DFA plot.

  7. BONUS POINT: Surround the points corresponding to each taxon with a polygon to ease visualization. Hint: the chull() function can be used to compute convex hulls, which is the name for polygons that enclose groups of points.