test_kernel_fits: Test multiple kernel fits to dispersal data.
In bochocki/correlatedtraits: Data and Analyses for [paper title]

test_kernel_fits takes output from wrangle_beetle_data and runs tests to determine the best-fitting dispersal kernel for all beetles as a whole, for females independently, and for males independently. The following kernels are tested: the Poisson distribution (Poisson), the negative binomial distribution (NegBinomial), the Sichel distribution (SICHEL), and the Poisson Inverse Gaussian distribution (a special case of the Sichel distribution where ν = -0.5).

1	test_kernel_fits(clean_data)

A list of lists. Each top-level list ($all, $females, or $males) contains the following objects:

$dist - A vector of the distances that were measured for each individual in the data subset.
$x - A vector of patch distances relevant to the study (x = 0:40).
$AIC - An AIC table ranking each distribution by its AIC score when fit to the data in $dist.
$frequency - The frequency of each distance measured in $dist, measured over the patches in $x.
$plot - A ggplot object showing the data in $frequency along with the best-fit plot of each candidate distribution.
$<distribution> - The results of maximum-likelihood fits of each distribution to the data in $dist, where <distribution> is either poisson, nbinom, pig, or sichel. The structure of these results exaclty follows the output from optim. For example, to get parameter estimates for fits to the Poisson distribution, using data from female beetles, use output$females$poisson$par.
$<distribution>$fit - The mass density function for a given distribution over $x (i.e., the line drawn in $plot.