Run an interpretation¶

Initialize a client with the server credentials and store them in the variable dai.

In [1]:

import driverlessai

dai = driverlessai.Client(address='http://localhost:12345', username='py', password='py')

import driverlessai dai = driverlessai.Client(address='http://localhost:12345', username='py', password='py')

Loading Data¶

Import the file creditcard.csv from S3 to the Driverless AI server and name the dataset 'creditcard-mli-interpretation'.

In [2]:

ds = dai.datasets.create(
    data='s3://h2o-public-test-data/mli-testing/manual-test/small-dataset/binomial/creditcard.csv', 
    data_source='s3', 
    name='creditcard-mli-interpretation'
)

ds = dai.datasets.create( data='s3://h2o-public-test-data/mli-testing/manual-test/small-dataset/binomial/creditcard.csv', data_source='s3', name='creditcard-mli-interpretation' )

Complete 100.00% - [4/4] Computed stats for column default payment next month

Create an experiment¶

Create a binomial classification experiment using Driverless AI.

The dai.experiments.create() function waits for the experiment to complete, and then returns an experiment object that can be used to run an interpretation.

In [3]:

ex = dai.experiments.create(train_dataset=ds,
                            target_column='default payment next month', 
                            task='classification',  
                            accuracy=1,
                            time=1,
                            interpretability=1,
                            name='creditcard-mli-interpretation')

ex = dai.experiments.create(train_dataset=ds, target_column='default payment next month', task='classification', accuracy=1, time=1, interpretability=1, name='creditcard-mli-interpretation')

Experiment launched at: http://localhost:12345/#/experiment?key=8810c64a-bf61-11ed-863e-d83bbfdbe6ce
Complete 100.00% - Status: Complete                                             

Create Interpretation¶

Interpret the experiment using the default explainers.

In [4]:

interpretation = dai.mli.create(
    experiment=ex,
    name="creditcard-mli-interpretation"
)

interpretation = dai.mli.create( experiment=ex, name="creditcard-mli-interpretation" )

Complete 100.00% - Interpretation successfully finished.

Explainer graphs must be viewed through the GUI.

In [5]:

interpretation.gui()

interpretation.gui()

Out[5]:

http://localhost:12345/#/interpret_next?interpret_key=dfbf31a6-bf61-11ed-863e-d83bbfdbe6ce

In [ ]: