Exam DP-100 topic 5 question 2 discussion

The correct answer is: 1. client = ExplanationClient.from_run_id() (from_run requires just one parameter and three are being passed here) 2. explanation = client.download_model_explanation() 3. explanation.get_feature_importance_dict() As dev2dev pointed out: https://github.com/MicrosoftDocs/azure-docs/blob/master/articles/machine-learning/how-to-machine-learning-interpretability-automl.md Check out in addition: https://docs.microsoft.com/es-es/python/api/azureml-interpret/azureml.interpret.explanationclient?view=azure-ml-py

client = ExplanationClient.from_run_id()# as run_id is passed as parmaeter explanation = client.download_model_explanation()# as it is mentioned in question it has #been uploaded feature_importances = explanation .get_feature_importance_dict()

client = ExplanationClient.from_run_id(workspace=ws, experiment_name=experiment.experiment_name, run_id=run.id) explanation = client.download_model_explanation() feature_importances = explanation.get_feature_importance_dict()

On exam 18-06-22

on exam 18/03/2022

on Exam 6 Nov 2021, the variable should be "client" in the first statement. However, the exam gives exactly the same typo.

after reading all the comments there is confusion in from_run and froun_run_id (correct answer)so if you follow the below link ExplanationClient.from_run(run) , it is expecting only 1 parameter and it is used to get an Explanation Client and upload the explanation whereas from_run_id is expecting 3 parameter and used to download the explanation from azureml.contrib.interpret.explanation.explanation_client import ExplanationClient client = ExplanationClient.from_run_id(workspace=ws, experiment_name=experiment.experiment_name, run_id=run.id) explanation = client.download_model_explanation() feature_importances = explanation.get_feature_importance_dict()

Agree with other posters' answers. Link for reference: https://docs.microsoft.com/en-us/learn/modules/explain-machine-learning-models-with-azure-machine-learning/4-create-explanations (Viewing the explanation section)

from azureml.interpret import ExplanationClient # Get the feature explanations client = ExplanationClient.from_run(run) engineered_explanations = client.download_model_explanation() feature_importances = engineered_explanations.get_feature_importance_dict() # Overall feature importance print('Feature\tImportance') for key, value in feature_importances.items(): print(key, '\t', value)

from_run_id(workspace, experiment_name, run_id)

Correct answer: from azureml.interpret import ExplanationClient # Get the feature explanations client = ExplanationClient.from_run(run) engineered_explanations = client.download_model_explanation() feature_importances = engineered_explanations.get_feature_importance_dict()

client = ExplanationClient.from_run_id(workspace=ws, experiment_name=experiment.experiment_name, run_id=run.id) explanation = client.download_model_explanation() feature_importances = explanation.get_feature_importance_dict()

The correct answer is : client = ExplanationClient.from_run(run) explanation = client.download_model_explanation() feature_importances = explanation .get_feature_importance_dict()

1 and 2 should be corrected Full example here: https://github.com/MicrosoftDocs/azure-docs/blob/master/articles/machine-learning/how-to-machine-learning-interpretability-automl.md from azureml.interpret import ExplanationClient client = ExplanationClient.from_run(best_run) raw_explanations = client.download_model_explanation(raw=True) print(raw_explanations.get_feature_importance_dict())

1. from_run() 2. client.download_model_explanation() 3. explanation.get_feature_importance_dict()

I think this solution is not correct the correct one is: 1. client = ExplanationClient.from_run_id() 2. explanation = client.download_model_explanation() 3. explanation.get_feature_importance_dict()

I think this is incorrect, 1 should be client = Explanation.from_run_id() 2. explanation = client.download_model_explanation() 3. explanation.get_feature_importance_dict()