phitter.simulation.process_simulation package

Submodules

phitter.simulation.process_simulation.process_simulation module

class phitter.simulation.process_simulation.process_simulation.ProcessSimulation

Bases: object

add_process(prob_distribution, parameters, process_id, number_of_products=1, number_of_servers=1, new_branch=False, previous_ids=None)

Add element to the simulation

Parameters:

  • prob_distribution (str) – Probability distribution to be used. You can use one of the following: ‘alpha’, ‘arcsine’, ‘argus’, ‘beta’, ‘beta_prime’, ‘beta_prime_4p’, ‘bradford’, ‘burr’, ‘burr_4p’, ‘cauchy’, ‘chi_square’, ‘chi_square_3p’, ‘dagum’, ‘dagum_4p’, ‘erlang’, ‘erlang_3p’, ‘error_function’, ‘exponential’, ‘exponential_2p’, ‘f’, ‘fatigue_life’, ‘folded_normal’, ‘frechet’, ‘f_4p’, ‘gamma’, ‘gamma_3p’, ‘generalized_extreme_value’, ‘generalized_gamma’, ‘generalized_gamma_4p’, ‘generalized_logistic’, ‘generalized_normal’, ‘generalized_pareto’, ‘gibrat’, ‘gumbel_left’, ‘gumbel_right’, ‘half_normal’, ‘hyperbolic_secant’, ‘inverse_gamma’, ‘inverse_gamma_3p’, ‘inverse_gaussian’, ‘inverse_gaussian_3p’, ‘johnson_sb’, ‘johnson_su’, ‘kumaraswamy’, ‘laplace’, ‘levy’, ‘loggamma’, ‘logistic’, ‘loglogistic’, ‘loglogistic_3p’, ‘lognormal’, ‘maxwell’, ‘moyal’, ‘nakagami’, ‘non_central_chi_square’, ‘non_central_f’, ‘non_central_t_student’, ‘normal’, ‘pareto_first_kind’, ‘pareto_second_kind’, ‘pert’, ‘power_function’, ‘rayleigh’, ‘reciprocal’, ‘rice’, ‘semicircular’, ‘trapezoidal’, ‘triangular’, ‘t_student’, ‘t_student_3p’, ‘uniform’, ‘weibull’, ‘weibull_3p’, ‘bernoulli’, ‘binomial’, ‘geometric’, ‘hypergeometric’, ‘logarithmic’, ‘negative_binomial’, ‘poisson’.

  • parameters (dict) – Parameters of the probability distribution.

  • process_id (str) – Unique name of the process to be simulated

  • number_of_products (int ,optional) – Number of elements that are needed to simulate in that stage. Value has to be greater than 0. Defaults equals to 1.

  • number_of_servers (int, optional) – Number of servers that process has and are needed to simulate in that stage. Value has to be greater than 0. Defaults equals to 1.

  • new_branch (bool ,optional) – Required if you want to start a new process that does not have previous processes. You cannot use this parameter at the same time with “previous_id”. Defaults to False.

  • previous_id (list[str], optional) – Required if you have previous processes that are before this process. You cannot use this parameter at the same time with “new_branch”. Defaults to None.

Return type:

None

process_graph(graph_direction='LR', save_graph_pdf=False)

Generates the graph of the process

Parameters:

  • graph_direction (str, optional) – You can show the graph in two ways: ‘LR’ left to right OR ‘TB’ top to bottom. Defaults to ‘LR’.

  • save_graph_pdf (bool, optional) – You can save the process graph in a PDF file. Defaults to False.

Return type:

None

run(number_of_simulations=1)

Simulation of the described process

Parameters:

number_of_simulations (int, optional) – Number of simulations of the process that you want to do. Defaults to 1.

Returns:

Results of every simulation requested

Return type:

list[float]

run_confidence_interval(confidence_level=0.95, number_of_simulations=1, replications=30)

Generates a confidence interval for the replications of the requested number of simulations.

Parameters:

  • confidence_level (float, optional) – Confidence required of the interval. Defaults to 0.95.

  • number_of_simulations (int, optional) – Number of simulations that are going to be run in each replication. Defaults to 1.

  • replications (int, optional) – Number of samples needed. Defaults to 30.

Returns:

Returns the lower bound, average, upper bound and standard deviation of the confidence interval

Return type:

tuple[float]

simulation_metrics()

Here you can find the average time per process and standard deviation

Returns:

Average and Standard deviation

Return type:

pd.DataFrame

Module contents