RobustAM
Purpose
RobustAM extends AMMCMC with heavy-tail and global-jump proposal components.
Full Sampling Section Keys
Sampling.Method(required): must beRobustAM.Sampling.Variables(required, array):name(required)description(required)distribution.type(required)distribution.parameters(required)- runtime-safe parameter sets:
Flat:min,maxLog:min,maxNormal:mean,stddevLog-Normal:mean,stddevLogit:location,scale
Sampling.LogLikelihood(required):{name, expression}list.Sampling.selection(optional, string)Sampling.Bounds:num_chains(required, integer)num_iters(required, integer)proposal_scale(required, number)adapt_enabled(optional, boolean, defaulttrue)adapt_start_iter(optional, integer, default100)adapt_window(optional, integer, default25)adapt_eps(optional, number, default1e-6)adapt_scale(optional, number, default2.38)global_jump_prob(optional, number, default0.08)heavy_tail_df(optional, number, default5.0)global_scale(optional, number, default1.8)
Full Skeleton
Sampling:
Method: "RobustAM"
Variables:
- name: p1
description: parameter 1
distribution:
type: Flat
parameters:
min: -5
max: 5
LogLikelihood:
- name: L_total
expression: "-0.5*(p1/1.0)^2"
Bounds:
num_chains: 8
num_iters: 10000
proposal_scale: 0.10
adapt_enabled: true
adapt_start_iter: 100
adapt_window: 25
adapt_eps: 1.0e-6
adapt_scale: 2.38
global_jump_prob: 0.10
heavy_tail_df: 4.0
global_scale: 2.0
Example
Sampling:
Method: "RobustAM"
Variables:
- name: m0
description: scalar mass
distribution:
type: Flat
parameters:
min: 100
max: 3000
LogLikelihood:
- name: L_total
expression: "L_higgs + L_dm"
Bounds:
num_chains: 8
num_iters: 10000
proposal_scale: 0.10
adapt_enabled: true
global_jump_prob: 0.10
heavy_tail_df: 4.0
global_scale: 2.0