rand_thompson: Simulate Bandit Problem (Thompson Sampling)
In dfcorbin/npbanditC: (Contextual) Multi-Armed Bandit Simulations

This function is not intended to be used on its own. Instead, use the higher level function 'sim_rand_bandit()'.

rand_thompson(
  nsteps,
  posteriors,
  basis,
  sd,
  J_mod,
  J_true,
  beta_true,
  calc_mse
)

`nsteps`	The number of steps in the trajectory.
`posteriors`	Posterior distributions must be passed as a single list, containing elements which are lists associated with each posterior distribution i.e. a list of lists. Each posterior distribution must contain "beta","covar","a","b". NOTE: EVEN IF THERE IS JUST ONE ELEMENT, THE INPUT MUST BE A LIST OF LISTS!
`basis`	A list of basis functions.
`sd`	A vector containing the standard deviation for each lever.
`J_mod`	The number of basis functions to use in the model (INCLUDING THE INTERCEPT).
`J_true`	The number of basis functions to use in thE TRUE expected reward functions (INCLUDING THE INTERCEPT).
`beta_true`	A matrix, where each row is a set of coefficients used to compute the true expected reward of each lever.