calculate_parbounds(dist, x; δ_cp = 0.1 )

compute the values at quantiles ±δcp around x with δcp difference in the cumulated probability. The quantiles are constrained to not extend beyond min_quant and max_quant.

It returns a NamedTuple of

• sens_lowe: lower quantile
• sens_upper: upper quantile
• cp_ref: cdf of x (cumulative distribution function, i.e. p-value)
• cpsenslower: cdf of lower quantile
• cpsensupper: cdf of upper quantile

A wider distribution prior distribution will result in a wider intervals.

The DataFrame variant assumes x as column :ref to be present and adds/modifies output columns (named as above outputs).

get matrix (nsample, npar) with uniformly sampled in cumaltive p domain