0

I have using pymc successfully, I believe. However, I would like to be able to visualize or plot a prior disctribution

em0 = pymc.Normal('em0',mu=emLog, tau=1./0.3, value=emLog)

where emLog is my guess at where to find the center of the distribution

em0 then has a number of methods but none seem to allow me to plot the distribution.

merv
  • 67,214
  • 13
  • 180
  • 245
Ken Dere
  • 1
  • 1
  • 1
  • I looked a little harder at the pymc (v2?) documentation and they had the code to do just this – Ken Dere Aug 18 '19 at 16:08
  • 1
    Is this using `pymc` or `pymc3`? I would recommend using `pymc3`, which has a function `sample_prior_predictive`. A more complete example would also be helpful -- in particular, you probably should not be using `emLog` as both the mean and the observed value of a parameter: typically there would be some unknown mean you would want to estimate _after_ observing `emLog` – colcarroll Aug 18 '19 at 16:09
  • More specifically: - what statement do you use to import `pymc` above? - is `emLog` a number, an array, or another random variable? – colcarroll Aug 18 '19 at 16:10
  • I am using pymc for now. Hope to get to pymc3 when I finish this. – Ken Dere Aug 19 '19 at 14:16
  • I just import pymc – Ken Dere Aug 19 '19 at 14:22

0 Answers0