Quick start guide

Here, we provide a basic example of importing data and running the inference model. More complex examples are available on the example notebooks page.

1. Fill out proxy data and age constraint tables according to the data formatting specifications.

2. Pre-process your data for model construction with load_data()

   from stratmc.data import load_data
   
   sample_data, age_data = load_data('path_to_proxy_data.csv', 'path_to_age_data.csv')
   

3. Build a pymc.model.core.Model with build_model(sample_data, age_data, proxies = ['proxy'])

   from stratmc.model import build_model
   
   model, gp = build_model(sample_data, age_data, proxies = ['d13c'])
   

4. Sample the model posterior using a JAX-assisted MCMC sampling algorithm with get_trace(model, gp, ages).

   import pymc as pm
   from stratmc.inference import get_trace
   
   # array of ages at which to sample the posterior proxy curve
   predict_ages = np.linspace(lower_age, upper_age, number_ages)
   
   trace = get_trace(model, gp, predict_ages)
   

5. Plot and analyze the results with the stratmc.plotting library.

   from stratmc import plotting
   
   plotting.proxy_inference(sample_data, age_data, trace)
   

   (Source code, png, hires.png, pdf)