ffp 0.2.2

Added relative_entropy() to compute the relative entropy between two distributions;
entropy-pooling() runs about to 30-60% faster (#19);
view_on_marginal_distribution() restricts just the first two moments of the distribution;
Small changes in the documentation files.

ffp 0.2.1

Bug fix in entropy_pooling() when dealing with multiple inequalities constraints;
Small changes in the documentation files.

A new class (ffp_views) was added to help the user to input views on the market for portfolio construction and simulation purposes. See the view_*() family of functions;
entropy_pooling() is now exported and can be used with three different solvers (nlminb, solnl, and nloptr);
bind_views() allows the user to “glue” multiple views into the same object;
ffp_moments() computes the location and dispersion parameters under flexible probabilities.

bind_probs() now returns the user call in the third column instead the old arbitrary key column (#13);

The package now is built upon S3 vctrs, instead of tibbles;
Functions bootstrap_scenarios(), scenario_density() and bind_probs() only accepts objects of ffp class;
Added ffp method to ggplot2::autoplot();
Exponential Smoothing is now computed with exp_decay().