Gyan's Data Viz

So I typically start any sort of project by just doing some data visualizations. I worked up a couple for you Gyan of the latest data file you sent me. I looked back at the model file I had sent you Gyan, because models won't really tell you much beyond what you can see and understand. So the first thing I did was look at effects of species and prey size. I plotted them together to show you what I found. These are just plots of the probability of successfully capturing prey with a binomial confidence interval.

plot of chunk unnamed-chunk-2

I've plotted them in size class groups with small = 1, med = 2, large = 3. The big thing here is to see that you've got pretty clear additive effects. baeza is clearly the best of all the species at capture effeciency. Also there's a strong size effect with large prey clearly the hardest. It looks like baeza's advantage goes away at large sizes.

Now let's look at effects of colony size. I just used logistic regression style plots with log web volume on the x-axis. One is coloured by size, the other by species.

plot of chunk unnamed-chunk-3

Here's the plot coloured by species. Clearly no effect, and no real differences by species within that lack of effect.

plot of chunk unnamed-chunk-4

pretty much the same for prey size. Clearly colony size doesn't seem to be having any effect and will probably come out as insignificant in any model.