Define model

You can define your model by declaring a new function. The function should take the data as input and return the predicted values.

Python
R

def model(Y, X):
    # Define the model here
    a = m.dist.normal(0, 1, name = 'a')
    b = m.dist.normal(1, 1, name = 'b')
    s = m.dist.exponential(1 name = 's')
    
    # Return the predicted values
    m.dist.normal(a + b * X, obs = Y)

model <- function(Y, X){
  # Define the model here
    a = m.dist.normal(0, 1, name = 'a')
    b = m.dist.normal(1, 1, name = 'b')
    s = m.dist.exponential(1 name = 's')
    m.dist.normal(a + b * X, obs = Y)
}

Note the additional obs argument when declaring the likelihood function. This argument is used to specify the observed data.

Other Formats