Estimated fixed coefficients in linear regression


#1

Hi,
I have a liner regression model where I do not want priors or samples for some of the coefficients (fixed coefficients).
I’d like these coefficients to be fixed and to see their value after I run the model (just like in a nonbayesian regression).
Is this possible in pymc3?
Thanks


#2

For small simple model, you can use the find_MAP() with Flat prior, which will give you maximum likelihood estimation (a point estimate).