train_pipeline.puml

@startuml
@startuml
skinparam backgroundColor #EEEBDC
skinparam handwritten true

skinparam sequence {
ArrowColor DeepSkyBlue
ActorBorderColor DeepSkyBlue
LifeLineBorderColor blue
LifeLineBackgroundColor #A9DCDF

ParticipantBorderColor DeepSkyBlue
ParticipantBackgroundColor DodgerBlue
ParticipantFontName Impact
ParticipantFontSize 17
ParticipantFontColor #A9DCDF

ActorBackgroundColor aqua
ActorFontColor DeepSkyBlue
ActorFontSize 17
ActorFontName Aapex
}

participant "main" as main
participant "train" as train
participant "collect_dataset" as collect_dataset
participant "model" as model
participant "test_dataset" as test_dataset
participant "Config" as Config
participant "run_cmd" as run_cmd
activate main
activate train
main -> train: if not args.skip_collect_dataset:
train -> collect_dataset
activate collect_dataset
collect_dataset -> run_cmd: external call 02-filter-and-move.py via make

collect_dataset -> make_patterns
make_patterns -> run_cmd: external call 04-find-patterns.py via make
activate Config
make_patterns -> Config: dataset_file()
Config -> make_patterns: file for saving results of 04-find-patterns.py
deactivate Config

collect_dataset -> run_cmd: external call 06-calculate-halstead.py via make
collect_dataset -> run_cmd: external call 07-merge.py via make
collect_dataset -> run_cmd: external call 08-split.py via make

deactivate collect_dataset
|||
main -> train_process
activate train_process
train_process -> train_process: read <train> dataset, made by 08-split.py
train_process -> train_process: scale_dataset
activate model
train_process -> model: fit_regressor()
deactivate model
train_process -> train_process: save model to file
activate model
train_process -> train_process: load model to file
train_process -> train_process: scale <test> dataset, made by 08-split.py
activate train_process  #DarkSalmon
train_process -> model: rank
note right : Smoke tests
model -> train_process: preds, importances
deactivate model
train_process -> train_process: remove trash catboost folder
deactivate train_process
train_process -> train
train -> main
deactivate train
deactivate main
@enduml