1

I am working on Parameter estimation using Metropolis-Hasting algorithm in Python (I am very new to Python). I have around 8 input parameters to estimate.

I referred to the PyMC documentation, the first step is to define a 'model' and pass it to the MCMC() method. How do I go about creating a 'model' in PyMC assuming a normal distribution for each parameter? Also, the values for the parameters need to be varied only by +\- 50% of their default values, How can I specify this in the model? should this be the standard deviation for that variable?

Any help is appreciated, Thanks!

user3453636
  • 77
  • 1
  • 9
  • It sounds like you're using PyMC -- if you are starting out, I would use PyMC3 instead, which is the current, maintained version of the project. You might also get better responses if you posted what you had tried so far! – colcarroll Jul 13 '17 at 14:55
  • are u still interested by an example ? – Stéphane Aug 04 '17 at 08:29

0 Answers0