dvc.yaml

stages:
  prepare:
    cmd: python3 src/data/prepare.py -path_train=data/raw/titanic.csv
    deps:
    - data/raw/titanic.csv
    - src/data/prepare.py
    outs:
    - data/processed/x_test.csv
    - data/processed/x_train.csv
    - data/processed/y_test.csv
    - data/processed/y_train.csv
  train:
    cmd: python3 src/models/train.py -x_train=data/processed/x_train.csv -y_train=data/processed/y_train.csv
      -path_pkl=models/model.pkl
    deps:
    - data/processed/x_train.csv
    - data/processed/y_train.csv
    - src/models/train.py
    params:
    - train_logistic.random_state
    - train_logistic.solver
    - train_logistic.penalty
    outs:
    - models/model.pkl
  evaluate:
    cmd: python3 src/models/evaluate.py -x_test=data/processed/x_test.csv -y_test=data/processed/y_test.csv
      -path_pkl=models/model.pkl -scores=reports/scores.json -plot=reports/plot.json
    deps:
    - data/processed/x_test.csv
    - data/processed/y_test.csv
    - models/model.pkl
    - src/models/evaluate.py
    metrics:
    - reports/scores.json:
        cache: false
    plots:
    - reports/plot.json:
        cache: false
        x: False_Positive_Rate
        y: True_Positive_Rate
params:
- dvclive/params.yaml
metrics:
- dvclive/metrics.json
plots:
- dvclive/plots/metrics:
    x: step