I became interested in pymc2's truncated pareto distribution for some random (no pun intended) reason.
In the context of checking for its implementation here, I found that the doc string seems to provide the wrong equation for the pdf (it seems the fortran code is actually doing the right thing after all though).
I think it should read,
f(x \mid \alpha, m, b) = \frac{\alpha m^{\alpha} x^{-\alpha - 1}}{1-(m/b)**{\alpha}}
instead of,
f(x \mid \alpha, m, b) = \frac{\alpha m^{\alpha} x^{-\alpha}}{1-(m/b)**{\alpha}}
I became interested in pymc2's truncated pareto distribution for some random (no pun intended) reason.
In the context of checking for its implementation here, I found that the doc string seems to provide the wrong equation for the pdf (it seems the fortran code is actually doing the right thing after all though).
I think it should read, f(x \mid \alpha, m, b) = \frac{\alpha m^{\alpha} x^{-\alpha - 1}}{1-(m/b)**{\alpha}}
instead of, f(x \mid \alpha, m, b) = \frac{\alpha m^{\alpha} x^{-\alpha}}{1-(m/b)**{\alpha}}
in the docstring.
As per here.