From: Bayesian Models for Astrophysical Data, Cambridge Univ. Press

(c) 2017,  Joseph M. Hilbe, Rafael S. de Souza and Emille E. O. Ishida 

 

you are kindly asked to include the complete citation if you used this material in a publication

​

​

​

​

Code 10.19 Bernoulli logit model, in Python using Stan, for assessing the relationship between Seyfert AGN activity and galactocentric distance

==================================================================================

import numpy as np
import pandas as pd
import pystan 
import statsmodels.api as sm

​

# Data
path_to_data = 'https://raw.githubusercontent.com/astrobayes/BMAD/master/data/Section_10p8/Seyfert.csv'

​

# read data
data_frame = dict(pd.read_csv(path_to_data))

​

x1 = data_frame['logM200']
x2 = data_frame['r_r200']

​

data = {}
data['Y'] = data_frame['bpt']
data['X'] = sm.add_constant(np.column_stack((x1,x2)))
data['K'] = data['X'].shape[1]
data['N'] = data['X'].shape[0]
data['gal'] = [0 if item == data_frame['zoo'][0] else 1 
                        for item in data_frame['zoo']]
data['P'] = 2

# Fit
stan_code="""
data{
    int<lower=0> N;                                                      # number of data points
    int<lower=0> K;                                                      # number of coefficients
    int<lower=0> P;                                                       # number of populations
    matrix[N,K] X;                                                         # [logM200, galactocentric distance]
    int<lower=0, upper=1> Y[N];                                  # Seyfert 1/AGN 0
    int<lower=0, upper=1> gal[N];                                # elliptical 0/spiral 1
}
parameters{
    matrix[K,P] beta;
    real<lower=0> sigma;
    real mu;
}
model{
    vector[N] pi;

    for (i in 1:N) {
        if (gal[i] == gal[1]) pi[i] = dot_product(col(beta,1),X[i]);
        else pi[i] = dot_product(col(beta,2), X[i]);
    }

    # shared hyperpriors
    sigma ~ gamma(0.001, 0.001);
    mu ~ normal(0, 100);

​

    # priors and likelihood
    for (i in 1:K) {
        for (j in 1:P) beta[i,j] ~ normal(mu, sigma);
    }

​

    Y ~ bernoulli_logit(pi);
}
"""

​

# Run mcmc
fit = pystan.stan(model_code=stan_code, data=data, iter=60000, chains=3,
                           warmup=30000, thin=10, n_jobs=3)

​

# Output
print(fit)

==================================================================================