Pipelines¶

class getml.pipeline.Pipelines(data=None)[source]¶

Container which holds all pipelines associated with the currently running project. The container supports slicing and is sort- and filterable.

Example:

Show the first 10 pipelines belonging to the current project:

getml.project.pipelines[:10]

You can use nested list comprehensions to retrieve a scoring history for your project:

import matplotlib.pyplot as plt

hyperopt_scores = [(score.date_time, score.mae) for pipe in getml.project.pipelines
                      for score in pipe.scores["data_test"]
                      if "hyperopt" in pipe.tags]

fig, ax = plt.subplots()
ax.bar(*zip(*hyperopt_scores))

Methods

filter(conditional)

Filters the pipelines container.

sort(key[, descending])

Sorts the pipelines container.