Azure-docs: How to unwrap azure model and convert it to its original model?

[yiikou]

How to unwrap azure model and convert it to its original model?

For example, I fetch the best model which is XGBoostClassifier, but the type of model is automl.client.core.common.model_wrappers.PipelineWithYTransformations. Now I want to convert it to XGBoost object so that I can use xgboost.plot_importance to analyse the model.

---

Document Details

⚠ Do not edit this section. It is required for docs.microsoft.com ➟ GitHub issue linking.

• ID: 585fff5e-cadd-8967-8c09-63d8afcb53b3
• Version Independent ID: 95274303-b152-2ee8-113e-e0763beca7ed
• Content: Create automated ML experiments - Azure Machine Learning service
• Content Source: articles/machine-learning/service/how-to-configure-auto-train.md
• Service: machine-learning
• Sub-service: core
• GitHub Login: @nacharya1
• Microsoft Alias: nilesha

[YutongTie-MSFT]

@yiikou Thank you for your feedback, we will investigate it and get back to you soon.

[YutongTie-MSFT]

@nacharya1 Hi, from my end I don't think we can do this. Could you please help to check if there any plan from product end? Thank you very much.

[nacharya1]

@Sabina [e-mail] can you look into this please?

From: YutongTie-MSFT [e-mail]
Sent: Monday, July 1, 2019 9:04 AM
To: MicrosoftDocs/azure-docs [e-mail]
Cc: Nilesh Acharya [e-mail]; Mention [e-mail]
Subject: Re: [MicrosoftDocs/azure-docs] How to unwrap azure model and convert it to its original model? (#33966)

@nacharya1https://nam06.safelinks.protection.outlook.com/?url=https%3A%2F%2Fgithub.com%2Fnacharya1&data=02%7C01%7Cnilesha%40microsoft.com%7Ca009a68bf5244892b35908d6fe3dbbfe%7C72f988bf86f141af91ab2d7cd011db47%7C1%7C0%7C636975938431780130&sdata=COx%2FNLPenyAwpEyexure2yuOnQ4MM7VryF5Aqj7WWJ8%3D&reserved=0 Hi, from my end I don't think we can do this. Could you please help to check if there any plan from product end? Thank you very much.

—
You are receiving this because you were mentioned.
Reply to this email directly, view it on GitHubhttps://nam06.safelinks.protection.outlook.com/?url=https%3A%2F%2Fgithub.com%2FMicrosoftDocs%2Fazure-docs%2Fissues%2F33966%3Femail_source%3Dnotifications%26email_token%3DAJ65JT2BLG5EQBPLVLT4YULP5ITHDA5CNFSM4H3LOD32YY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGODY6TFAQ%23issuecomment-507327106&data=02%7C01%7Cnilesha%40microsoft.com%7Ca009a68bf5244892b35908d6fe3dbbfe%7C72f988bf86f141af91ab2d7cd011db47%7C1%7C0%7C636975938431790126&sdata=WbkCrTuoMF0q9o7y3Sw5AizkLC059NBIYwOi7JgXEfk%3D&reserved=0, or mute the threadhttps://nam06.safelinks.protection.outlook.com/?url=https%3A%2F%2Fgithub.com%2Fnotifications%2Funsubscribe-auth%2FAJ65JT5PGCM6E7V2DGRCQQTP5ITHDANCNFSM4H3LOD3Q&data=02%7C01%7Cnilesha%40microsoft.com%7Ca009a68bf5244892b35908d6fe3dbbfe%7C72f988bf86f141af91ab2d7cd011db47%7C1%7C0%7C636975938431790126&sdata=ouCD%2Bx7nf1MS8cOlNBN8%2FHaYqsqUpm9XGm1oNj7Gkmo%3D&reserved=0.

[marabout2015]

Hello @yiikou,

What do you see when you type:

automl_eclipse_2_0_xgboost.named_steps["XGBoostClassifier"].plot_importance()

Mario

[yiikou]

hi @marabout2015 , it will cause error, since plot_importance is from the xgboost package.

[j-martens]

@marabout2015 @cartacioS - any other guidance for @yiikou please?

[ahyerman]

@yiikou you should be able to retrieve the estimator through automl_eclipse_2_0_xgboost.steps[-1][1] or as @marabout2015 mentioned through named steps automl_eclipse_2_0_xgboost.named_steps["XGBoostClassifier"]

The code would look like this:

model = automl_eclipse_2_0_xgboost.steps[-1][1]
xgboost.plot_importance(model)