[x] Make a scatterplot comparing the number of links sampled using the metaweb and local network methods (averaged across $n$ simulations), each point corresponding to a different network.
[x] Make a scatterplot (or line plot) of the distance (KL divergence) between the distribution of the number of links obtained from the local network and metaweb methods against the number of simulations. We should expect the difference to shrink with the number of simulations.
Spatial scaling figure
[x] Update the plot using a range of initial latitudes instead of a fixed one (sliding window).
[x] Make a quantile plot of the expected number of links in the metaweb (species subnetwork) as a function of latitude width. Remove the expected total number of links panel.
[x] Add data points to the curve.
[x] Make a heatmap of the expected number of links in the merged network (z) as a function of the latitude width (y) and position (x).
Spatiotemporal model figure
[x] Add the points of the $\lambda$ and $\gamma$ parameters used to make the line plots to the heatmap.
For the network sampling figure, I decided to plot connectance instead of the number of links and use the mean squared logarithmic error as my measure of distance.
Network accumulation figure
Nothing to change 👍
Network sampling figure
Spatial scaling figure
Spatiotemporal model figure