python - pymc3 Estimated Parameter Constraint -


i trying estimate matrix assumed exponentially distributed. however, add l1-norm constraint columns of matrix. example, matrix estimated line given below , want l1-norm of each w column equals 1. in other words, sum(w[:,1])=sum(w[:,2])=...sum(w[:,ns])=1.

the estimation line here: w = pm.exponential('w_act', lam=theta_w, shape=(nc, ns))

do have suggestions problem?


Comments

Popular posts from this blog

python - Selenium remoteWebDriver (& SauceLabs) Firefox moseMoveTo action exception -

Ansible warning on jinja2 braces on when -

html - How to custom Bootstrap grid height? -