Two Group - Allow users to specify Regimes instead of using a predetermined config.

[IsmailM]

See:

hygeia/TwoGroup/run_inference_two_groups.py

Lines 20 to 23 in 903d8fc

	flags.DEFINE_integer("regimes_config",
	default=8,
	help="regime configuration id for mean/stds of regime params")
	flags.DEFINE_integer("minimum_duration",

hygeia/TwoGroup/run_inference_two_groups.py

Lines 100 to 184 in 903d8fc

	###############################################################################
	## PARAMETERS
	###############################################################################
	## The following parameters are fixed/assumed to be known, i.e.
	## they are not estimated by the parameter-estimation scheme.
	## They are also the same for both the case and control groups:
	if FLAGS.regimes_config == 0:
	# regime specific mean parameters beta laws
	mu_true = tf.Variable([0.99, 0.01, 0.80, 0.20, 0.50, 0.50],
	dtype = dtype, name = 'mu_true')
	# regime specific standard-deviation parameters beta laws
	sigma_true = tf.Variable(np.array([0.05, 0.05, 0.20, 0.20, 0.20, 1 / np.sqrt(12)]),
	dtype = dtype, name = 'sigma_true')
	elif FLAGS.regimes_config == 1:
	# regime specific mean parameters beta laws
	mu_true = tf.Variable([0.95, 0.05, 0.85, 0.15, 0.50, 0.50],
	dtype = dtype, name = 'mu_true')
	# regime specific standard-deviation parameters beta laws
	sigma_true = tf.Variable(np.array([0.10, 0.10, 0.08, 0.08, 0.20, 1 / np.sqrt(12)]),
	dtype = dtype, name = 'sigma_true')
	elif FLAGS.regimes_config == 2:
	# regime specific mean parameters beta laws
	mu_true = tf.Variable([0.95, 0.05, 0.85, 0.15, 0.50, 0.50],
	dtype = dtype, name = 'mu_true')
	# regime specific standard-deviation parameters beta laws
	sigma_true = tf.Variable(np.array([0.10, 0.10, 0.04, 0.04, 0.20, 1 / np.sqrt(12)]),
	dtype = dtype, name = 'sigma_true')
	elif FLAGS.regimes_config == 3:
	# regime specific mean parameters beta laws
	mu_true = tf.Variable([0.95, 0.05, 0.85, 0.15, 0.50, 0.50],
	dtype = dtype, name = 'mu_true')
	# regime specific standard-deviation parameters beta laws
	sigma_true = tf.Variable(np.array([0.15, 0.15, 0.08, 0.08, 0.20, 1 / np.sqrt(12)]),
	dtype = dtype, name = 'sigma_true')
	elif FLAGS.regimes_config == 4:
	# regime specific mean parameters beta laws
	mu_true = tf.Variable([0.95, 0.05, 0.85, 0.15, 0.50, 0.50],
	dtype = dtype, name = 'mu_true')
	# regime specific standard-deviation parameters beta laws
	sigma_true = tf.Variable(np.array([0.15, 0.15, 0.04, 0.04, 0.20, 1/np.sqrt(12)]),
	dtype = dtype, name = 'sigma_true')
	elif FLAGS.regimes_config == 5:
	# regime specific mean parameters beta laws
	mu_true = tf.Variable([0.99, 0.01, 0.80, 0.20, 0.50, 0.50],
	dtype = dtype, name = 'mu_true')
	# regime specific standard-deviation parameters beta laws
	sigma_true = tf.Variable(np.array([0.05, 0.05, 0.15, 0.15, 0.20, 1 / np.sqrt(12)]),
	dtype = dtype, name = 'sigma_true')
	elif FLAGS.regimes_config == 6:
	# regime specific mean parameters beta laws
	mu_true = tf.Variable([0.95, 0.05, 0.85, 0.15, 0.50, 0.50],
	dtype = dtype, name = 'mu_true')
	# regime specific standard-deviation parameters beta laws
	sigma_true = tf.Variable(np.array([0.05, 0.05, 0.05, 0.05, 0.1, 1 / np.sqrt(12)]),
	dtype = dtype, name = 'sigma_true')
	elif FLAGS.regimes_config == 7:
	# regime specific mean parameters beta laws
	mu_true = tf.Variable([0.99, 0.01, 0.80, 0.20, 0.50, 0.50, 0.50],
	dtype = dtype, name = 'mu_true')
	# regime specific standard-deviation parameters beta laws
	sigma_true = tf.Variable(np.array([0.05, 0.05, 0.15, 0.15, 0.20, 1 / np.sqrt(12), 1 / np.sqrt(6)]),
	dtype = dtype, name = 'sigma_true')
	elif FLAGS.regimes_config == 8:
	# regime specific mean parameters beta laws
	mu_true = tf.Variable([0.95, 0.05, 0.80, 0.20, 0.50, 0.50],
	dtype = dtype, name = 'mu_true')
	# regime specific standard-deviation parameters beta laws
	sigma_true = tf.Variable(np.array([0.05, 0.05, 0.1, 0.1, 0.1, 1 / np.sqrt(12)]),
	dtype = dtype, name = 'sigma_true')
	elif FLAGS.regimes_config == 9:
	# regime specific mean parameters beta laws
	mu_true = tf.Variable([0.95, 0.05, 0.75, 0.25, 0.50, 0.50],
	dtype = dtype, name = 'mu_true')
	# regime specific standard-deviation parameters beta laws
	sigma_true = tf.Variable(np.array([0.1, 0.1, 0.1, 0.1, 0.1, 1 / np.sqrt(12)]),
	dtype = dtype, name = 'sigma_true')
	elif FLAGS.regimes_config == 10:
	# regime specific mean parameters beta laws
	mu_true = tf.Variable([0.95, 0.05, 0.80, 0.20, 0.50],
	dtype = dtype, name = 'mu_true')
	# regime specific standard-deviation parameters beta laws
	sigma_true = tf.Variable(np.array([0.05, 0.05, 0.1, 0.1, 0.1]),
	dtype = dtype, name = 'sigma_true')
	n_methylation_regimes = mu_true.shape[0]