JohannesBuchner / UltraNest

Fit and compare complex models reliably and rapidly. Advanced nested sampling.
https://johannesbuchner.github.io/UltraNest/
Other
142 stars 30 forks source link

Prior docs issues #69

Closed JohannesBuchner closed 2 years ago

JohannesBuchner commented 2 years ago

https://johannesbuchner.github.io/UltraNest/priors.html#Dependent-priors

1) last line before return is wrong:

def transform_correlated(quantiles):
    parameters = np.empty_like(quantiles)
    # first parameter is independent
    parameters[0] = gauss1.ppf(quantiles[0])
    # second parameter depends on first parameter, here with a shift
    parameters[1] = parameters[0] + gauss2.ppf(quantiles[0])
    return parameters

should be

def transform_correlated(quantiles):
    parameters = np.empty_like(quantiles)
    # first parameter is independent
    parameters[0] = gauss1.ppf(quantiles[0])
    # second parameter depends on first parameter, here with a shift
    parameters[1] = parameters[0] + gauss2.ppf(quantiles[1])
    return parameters

This is also not explained in a lot of detail.

2) There are some unnecessary print outs.

3) More math formulas could help.

JohannesBuchner commented 2 years ago

Fixed with the new docs and latest release (3.4.6).