Configuration-Driven Workflow

HIcosmo supports Cobaya-like YAML configuration.

Parallel Initialization

Before running any configuration, initialize the parallel environment:

import hicosmo as hc
hc.init(8)  # 8 parallel devices

Note

hc.init(N) creates N JAX logical devices, enabling true parallelism for N chains.

hc.init(8) - 8 parallel devices (recommended)
hc.init() - Auto-detect (up to 8)
hc.init("GPU") - GPU mode

YAML Configuration Example

name: joint_sn_bao
theory: LCDM
likelihood:
  - name: sn
    dataset: pantheon+
  - name: bao
    dataset: sdss_dr12

params:
  H0: {prior: {dist: uniform, min: 50, max: 100}, ref: 70.0}
  Omega_m: {prior: {dist: uniform, min: 0.1, max: 0.5}, ref: 0.3}

sampler:
  name: numpyro
  num_samples: 1000
  num_chains: 8         # Default = number of devices
  num_warmup: 300

output:
  root: results
  chain_name: joint_sn_bao

Running a Configuration

# 1. Initialize first
import hicosmo as hc
hc.init(8)

# 2. Run the configuration
from hicosmo import run_from_config
run_from_config("examples/tutorials/configs/joint_sn_bao.yaml")