plot.mob_out uses rounded sampling effort for plot-based rarefaction curve

[dmcglinn]

The function plot.mob_out provides the plot-based rarefaction curve which is the individual rarefaction curve evaluated at # of plots * treatment specific density. Currently this sampling effort is rounded (from the script ./R/mobr.R):

get_sample_curves  = function(...){
    ...
    nplots = nrow(comm_level)
    level_dens = sum(comm_level) / nplots
    samp_effort = round((1:nplots) * level_dens)
    impl_S = rarefaction(comm_level, 'indiv', samp_effort)
    ...
}

The need to round the sampling effort is because of the binomial coefficient only takes integers; however, the formula for individual-based rarefaction can be reformulated using the gamma function which can handle non-integer values. We could use instead:

    impl_S = rarefaction(comm_level, 'indiv', samp_effort, dens_ratio = 1 + 1e-14)

This would harness the gamma formulated form but not rescale the curve any. @rueuntal what do you think - probably this isn't going to make a huge deal but it seems like it would be nice to get away from rounding the sampling effort since we have the code in place to do this anyways.

[rueuntal]

Sounds good to me!