exports Module

Exports a pipeline in DOT language. Relies on method <xref:nimbusml.pipeline.Pipeline.get_fit_info>.

The function shows intermediate columns between operators. Blue columns are left unchanged, yellow columns are either created or replaced.

   import pandas
   from nimbusml.linear_model import FastLinearRegressor
   from nimbusml.feature_extraction.categorical import OneHotVectorizer
   from nimbusml.preprocessing.normalization import MeanVarianceScaler
   from nimbusml.preprocessing.schema import ColumnDropper
   from nimbusml import Pipeline
   from nimbusml.utils.exports import dot_export_pipeline

   df = pandas.DataFrame(dict(education=['A', 'B', 'A', 'B', 'A'],
                              workclass=['X', 'X', 'Y', 'Y', 'Y'],
                              yy=[1.1, 2.2, 1.24, 3.4, 3.4]))
   X = df.drop('yy', axis=1)
   y = df['yy']

   exp = Pipeline([
               MeanVarianceScaler() << {'new_y': 'yy'},
               OneHotVectorizer() << ['workclass', 'education'],
               ColumnDropper() << 'yy',
               FastLinearRegressor() << {'Feature': ['workclass',
                                                     'education'],
                                          Role.Label: 'new_y'}
               ])

   dot = dot_export_pipeline(exp, X, y)
   print(doc)

img_export_pipeline uses this function to render the graph as an image.

dot_export_pipeline(pipeline, X, y=None, **params)

Parameters

Produces an image which represents the data and the pipelines steps. It converts the export returned by function dot_export_pipeline and returns a graph built by module graphviz.

   import pandas
   from nimbusml.linear_model import FastLinearRegressor
   from nimbusml.feature_extraction.categorical import OneHotVectorizer
   from nimbusml.preprocessing.normalization import MeanVarianceScaler
   from nimbusml.preprocessing.schema import ColumnDropper
   from nimbusml import Pipeline
   from nimbusml.utils.exports import img_export_pipeline

   df = pandas.DataFrame(dict(education=['A', 'B', 'A', 'B', 'A'],
                              workclass=['X', 'X', 'Y', 'Y', 'Y'],
                              yy=[1.1, 2.2, 1.24, 3.4, 3.4]))
   X = df.drop('yy', axis=1)
   y = df['yy']

   exp = Pipeline([
               MeanVarianceScaler() << {'new_y': 'yy'},
               OneHotVectorizer() << ['workclass', 'education'],
               ColumnDropper() << 'yy',
               FastLinearRegressor() << {'Feature': ['workclass',
                                                     'education'],
                                         Role.Label: 'new_y'}
               ])

   img_export_pipeline(exp, X, y).render("mypipeline.png")

img_export_pipeline(pipeline, X, y=None, **params)

Parameters