Support fitting of Normal and other distributions

[aphalo]

As with other model fits, but in this case for univariate data, support plotting of prediction and annotation with parameter estimates.

• stat_edf_area(): ggplot2::stat_density() does the job for plotting, a wrapper supporting fewer kernels could be useful for consistency with stat_edf_params().

• stat_edf_params(): numeric values and character labels for the fitted distributions. ~~One option would be assuming normality and simply computing mean() and sd.~~

• Are these needed at all? ~~Probably not.~~ Fitting one Normal is just a special case of mixture of Normals.

• [x] Consider if the same stats would be able to support univariate mixtures. #85 Yes for Normal

• [x] Consider which annotations in addition to $\bar x$ and $\sigma$ would be useful. The SE of these parameter estimates.

• [ ] Can kurtosis parameters be estimated and made available as labels?

• [ ] What other distributions can be supported?

[aphalo]

The best bet for adding support for fitting various distributions individually seems to be to use MASS::fitdstr() as it is safe to depend on 'MASS' as is already being imported.

[aphalo]

stat_distrmix_line() and stat_distrmix_line() support also fitting a single distribution. Currently, the Normal.

• [ ] What other distributions can be supported?
• lognormal, possibly
• Gamma, possibly