python | how can modify EM algorithm to select random columns

I want to select a random set of columns from dataframe and then calculate probability, the best algorithm for probability is expectation–maximization (EM) algorithm, but with modification to select randomly. The final output is like that ( for every run, the algorithm selected random columns):

Feature    | likelihood 
 F1, F2, F3 |  56   
 F2, F4, F6 |  78 

the code:

import random

def em(data, columns, num_iterations):
  # Initialize the model parameters
  model_params = initialize_model_params(data, columns)

  for i in range(num_iterations):
    # Select a random subset of columns
    selected_columns = random.sample(columns, k=len(columns) // 2)

    # E step: calculate the expected value of the hidden variables using the selected columns
    expected_hidden_vars = calculate_expected_hidden_vars(data, selected_columns, model_params)

    # M step: update the estimates of the model parameters using the selected columns and the expected hidden variables
    model_params = update_model_params(data, selected_columns, expected_hidden_vars)

  # Return the maximum likelihood estimates of the model parameters
  return model_params

def initialize_model_params(data, columns):
  # Initialize the model parameters using the data in the columns
  # Return the initialized parameters

def calculate_expected_hidden_vars(data, columns, model_params):
  # Calculate the probabilities of the different values that the columns can take on
  # Use the current estimates of the model parameters and the probabilities to calculate the expected value of the hidden variables
  # Return the expected hidden variables

def update_model_params(data, columns, expected_hidden_vars):
  # Use the expected values of the hidden variables to update the estimates of the model parameters
  # Return the updated model parameters

how can definitions of the functions: initialize_model_params , calculate_expected_hidden_vars, update_model_params