# Deploy ## predict\_model This function generates the label using a trained model. When `data` is None, it predicts label and score on the holdout set. ### **Hold-out predictions** {% code lineNumbers="true" %} ```python # load dataset from pycaret.datasets import get_data diabetes = get_data('diabetes') # init setup from pycaret.classification import * clf1 = setup(data = diabetes, target = 'Class variable') # create a model xgboost = create_model('xgboost') # predict on hold-out predict_model(xgboost) ``` {% endcode %} ![Output from predict\_model(xgboost)](https://3811348345-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FjAq5m5T7Qtz03TnB0Wve%2Fuploads%2FxDoQCZokWwjWAoj6V3RG%2Fimage.png?alt=media\&token=b4b861cd-bf71-48fc-a059-20f276dd34f0) ### **Unseen data predictions** {% code lineNumbers="true" %} ```python # load dataset from pycaret.datasets import get_data diabetes = get_data('diabetes') # init setup from pycaret.classification import * clf1 = setup(data = diabetes, target = 'Class variable') # create a model xgboost = create_model('xgboost') # predict on new data new_data = diabetes.copy() new_data.drop('Class variable', axis = 1, inplace = True) predict_model(xgboost, data = new_data) ``` {% endcode %} ![Output from predict\_model(xgboost, data=new\_data)](https://3811348345-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FjAq5m5T7Qtz03TnB0Wve%2Fuploads%2Fc6GqPUXIf7G1WWmgTUHp%2Fimage.png?alt=media\&token=f0287182-98f1-498c-9f97-f09e87ea8c22) ### Probability by class {% hint style="info" %} **NOTE:** This is only applicable for the [Classification](https://pycaret.gitbook.io/docs/get-started/modules) use-cases. {% endhint %} {% code lineNumbers="true" %} ```python # load dataset from pycaret.datasets import get_data diabetes = get_data('diabetes') # init setup from pycaret.classification import * clf1 = setup(data = diabetes, target = 'Class variable') # create a model xgboost = create_model('xgboost') # predict on new data new_data = diabetes.copy() new_data.drop('Class variable', axis = 1, inplace = True) predict_model(xgboost, raw_score = True, data = new_data) ``` {% endcode %} ![Output from predict\_model(xgboost, raw\_score = True, data = new\_data)](https://3811348345-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FjAq5m5T7Qtz03TnB0Wve%2Fuploads%2FwXESCzmpMTomRPTLjPdS%2Fimage.png?alt=media\&token=349b4985-7396-41be-b622-968d0d94344c) ### Setting probability threshold {% hint style="info" %} **NOTE:** This is only applicable for the [Classification](https://pycaret.gitbook.io/docs/get-started/modules) use-cases (binary only). {% endhint %} The threshold for converting predicted probability to the class labels. Unless this parameter is set, it will default to the value set during model creation. If that wasn’t set, the default will be 0.5 for all classifiers. Only applicable for binary classification. {% code lineNumbers="true" %} ```python # load dataset from pycaret.datasets import get_data diabetes = get_data('diabetes') # init setup from pycaret.classification import * clf1 = setup(data = diabetes, target = 'Class variable') # create a model xgboost = create_model('xgboost') # probability threshold 0.3 predict_model(xgboost, probability_threshold = 0.3) ``` {% endcode %} ![Output from predict\_model(xgboost, probability\_threshold = 0.3)](https://3811348345-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FjAq5m5T7Qtz03TnB0Wve%2Fuploads%2FI18qJsoiBqbV28pJhBMT%2Fimage.png?alt=media\&token=efaa70ce-e67b-400c-8dd9-91bc9cf24359) #### Comparison between different thresholds on the hold-out data ![probability threshold = 0.5 vs. probability threshold = 0.3](https://3811348345-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FjAq5m5T7Qtz03TnB0Wve%2Fuploads%2FDlMx2jStEp9xEro1wTg4%2Fimage.png?alt=media\&token=cf7c9745-94ed-4f65-a4ab-9e31d42b0f41) ## finalize\_model This function trains a given model on the entire dataset including the hold-out set. {% code lineNumbers="true" %} ```python # load dataset from pycaret.datasets import get_data diabetes = get_data('diabetes') # init setup from pycaret.classification import * clf1 = setup(data = diabetes, target = 'Class variable') # create a model rf = create_model('rf') # finalize a model finalize_model(rf) ``` {% endcode %} ![Output from finalize\_model(rf)](https://3811348345-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FjAq5m5T7Qtz03TnB0Wve%2Fuploads%2Fh56NcrnyMOs3dIhuOrU5%2Fimage.png?alt=media\&token=fb783786-5568-4273-b50c-bbad7e915b82) This function doesn't change any parameter of the model. It only refits on the entire dataset including the hold-out set. ## deploy\_model This function deploys the entire ML pipeline on the cloud. {% code overflow="wrap" lineNumbers="true" %} ```python # load dataset from pycaret.datasets import get_data diabetes = get_data('diabetes') # init setup from pycaret.classification import * clf1 = setup(data = diabetes, target = 'Class variable') # create a model lr = create_model('lr') # finalize a model final_lr = finalize_model(lr) # deploy a model deploy_model(final_lr, model_name = 'lr_aws', platform = 'aws', authentication = { 'bucket' : 'pycaret-test' }) ``` {% endcode %} ![Output from deploy\_model(...)](https://3811348345-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FjAq5m5T7Qtz03TnB0Wve%2Fuploads%2FxQUVp2IUnC3QuFqu47Ib%2Fimage.png?alt=media\&token=b799b4af-d1a0-4a0a-9085-a22c2e966552) ### AWS Before deploying a model to an AWS S3 (‘aws’), environment variables must be configured using the command-line interface. To configure AWS environment variables, type **aws configure** in your python command line. The following information is required which can be generated using the Identity and Access Management (IAM) portal of your amazon console account: * AWS Access Key ID * AWS Secret Key Access * Default Region Name (can be seen under Global settings on your AWS console) * Default output format (must be left blank) ### GCP To deploy a model on Google Cloud Platform ('gcp'), the project must be created using the command-line or GCP console. Once the project is created, you must create a service account and download the service account key as a JSON file to set environment variables in your local environment. Learn more about it: ### Azure To deploy a model on Microsoft Azure ('azure'), environment variables for the connection string must be set in your local environment. Go to settings of storage account on Azure portal to access the connection string required. * AZURE\_STORAGE\_CONNECTION\_STRING (required as environment variable) Learn more about it: ## save\_model This function saves the transformation pipeline and a trained model object into the current working directory as a pickle file for later use. {% code lineNumbers="true" %} ```python # load dataset from pycaret.datasets import get_data diabetes = get_data('diabetes') # init setup from pycaret.classification import * clf1 = setup(data = diabetes, target = 'Class variable') # create a model dt = create_model('dt') # save pipeline save_model(dt, 'dt_pipeline') ``` {% endcode %} ![Output from save\_model(dt, 'dt\_pipeline')](https://3811348345-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FjAq5m5T7Qtz03TnB0Wve%2Fuploads%2Fh74jxGDC5EojHs9Xuwhs%2Fimage.png?alt=media\&token=d9a27a4e-521e-4b00-aed2-843176b86a7e) ## load\_model This function loads a previously saved pipeline. {% code lineNumbers="true" %} ```python # load dataset from pycaret.datasets import get_data diabetes = get_data('diabetes') # init setup from pycaret.classification import * clf1 = setup(data = diabetes, target = 'Class variable') # create a model dt = create_model('dt') # save pipeline save_model(dt, 'dt_pipeline') # load pipeline load_model('dt_pipeline') ``` {% endcode %} ![Output from load\_model('dt\_pipeline')](https://3811348345-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FjAq5m5T7Qtz03TnB0Wve%2Fuploads%2FdlwPXLiInZk52WdFaQ8D%2Fimage.png?alt=media\&token=597f91a9-5d83-40f9-b333-5e3a22e17aa2) ## save\_experiment The `save_experiment` function saves the experiment to a pickle file. The experiment is saved using cloudpickle to deal with lambda functions. The data or test data is NOT saved with the experiment and will need to be specified again when loading using `load_experiment`. {% code lineNumbers="true" %} ```python # load dataset from pycaret.datasets import get_data diabetes = get_data('diabetes') # init setup from pycaret.classification import * clf1 = setup(data = diabetes, target = 'Class variable') # save experiment save_experiment('my_saved_experiment1') ``` {% endcode %} ## load\_experiment The `load_experiment` function loads an experiment from the path or a file. The `data` (and `test_data`) is not saved with the experiment and will need to be specified again at the time of loading. {% code lineNumbers="true" %} ```python # load data data = get_data('diabetes') # load experiment function from pycaret.classification import load_experiment clf2 = load_experiment('my_saved_experiment1', data = data) ``` {% endcode %}
## check\_drift The `check_drift` function generates a drift report file using the evidently library. {% code overflow="wrap" lineNumbers="true" %} ```python # load dataset from pycaret.datasets import get_data data = get_data('insurance') # generate drift report check_drift(reference_data = data.head(500), current_data = data.tail(500), target = 'charges') ``` {% endcode %} It will generate a HTML report locally.
## convert\_model This function transpiles the trained machine learning model's decision function in different programming languages such as Python, C, Java, Go, C#, etc. It is very useful if you want to deploy models into environments where you can't install your normal Python stack to support model inference. {% code lineNumbers="true" %} ```python # load dataset from pycaret.datasets import get_data juice = get_data('juice') # init setup from pycaret.classification import * exp_name = setup(data = juice, target = 'Purchase') # train a model lr = create_model('lr') # convert a model convert_model(lr, 'java') ``` {% endcode %} ![Output from convert\_model(lr, 'java')](https://3811348345-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FjAq5m5T7Qtz03TnB0Wve%2Fuploads%2FUNuM0Ay7LLiCjDPNUdC5%2Fimage.png?alt=media\&token=551b0075-552f-44bb-9be6-4cc2c59be930) #### Video: {% embed url="" %} ## create\_api This function takes an input model and creates a POST API for inference. It only creates the API and doesn't run it automatically. To run the API, you must run the Python file using `!python`. {% code lineNumbers="true" %} ```python # load dataset from pycaret.datasets import get_data juice = get_data('juice') # init setup from pycaret.classification import * exp_name = setup(data = juice, target = 'Purchase') # train a model lr = create_model('lr') # create api create_api(lr, 'lr_api') # run api !python lr_api.py ``` {% endcode %} ![Output from create\_api(lr, 'lr\_api')](https://3811348345-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FjAq5m5T7Qtz03TnB0Wve%2Fuploads%2FChPOL5r6IcWgkj1OBm3h%2Fimage.png?alt=media\&token=83b701d8-1d3a-4ee1-a500-d7d11d4f7183) Once you initialize API with the `!python` command. You can see the server on localhost:8000/docs. ![FastAPI server hosted on localhost](https://3811348345-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FjAq5m5T7Qtz03TnB0Wve%2Fuploads%2FpzaQUHxf4Smt3390Mi2q%2Fimage.png?alt=media\&token=415d1363-b2c7-4e11-b2ae-80410682de23) #### Video: {% embed url="" %} ## create\_docker This function creates a `Dockerfile` and `requirements.txt` for productionalizing API end-point. {% code lineNumbers="true" %} ```python # load dataset from pycaret.datasets import get_data juice = get_data('juice') # init setup from pycaret.classification import * exp_name = setup(data = juice, target = 'Purchase') # train a model lr = create_model('lr') # create api create_api(lr, 'lr_api') # create docker create_docker('lr_api') ``` {% endcode %} ![Output from create\_docker('lr\_api')](https://3811348345-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FjAq5m5T7Qtz03TnB0Wve%2Fuploads%2F4Fpj68sWSQLSWcw5d7fr%2Fimage.png?alt=media\&token=6540627f-93ba-4877-8ef9-2a3ae3be7e6b) You can see two files are created for you. ![%load requirements.txt](https://3811348345-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FjAq5m5T7Qtz03TnB0Wve%2Fuploads%2FiND3URES8duiUpF8YAwF%2Fimage.png?alt=media\&token=569dad5c-f88f-4f8f-89a3-ac0a6c1c1a79) ![%load DockerFile](https://3811348345-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FjAq5m5T7Qtz03TnB0Wve%2Fuploads%2F2crfM6bIoe4u4CMIAcMo%2Fimage.png?alt=media\&token=36d8fbde-83cf-460c-a3b0-640b40d4dcb0) #### Video: {% embed url="" %} ## create\_app This function creates a basic `gradio` app for inference. It will later be expanded for other app types such `Streamlit`. {% code lineNumbers="true" %} ```python # load dataset from pycaret.datasets import get_data juice = get_data('juice') # init setup from pycaret.classification import * exp_name = setup(data = juice, target = 'Purchase') # train a model lr = create_model('lr') # create app create_app(lr) ``` {% endcode %} ![Output from create\_app(lr)](https://3811348345-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FjAq5m5T7Qtz03TnB0Wve%2Fuploads%2F39qYrZUTzgeJPLcfgMpA%2Fimage.png?alt=media\&token=88638e7a-689a-4a92-941f-07037379ee83) #### Video: {% embed url="" %} --- # Agent Instructions: Querying This Documentation If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question. Perform an HTTP GET request on the current page URL with the `ask` query parameter: ``` GET https://pycaret.gitbook.io/docs/get-started/functions/deploy.md?ask= ``` The question should be specific, self-contained, and written in natural language. The response will contain a direct answer to the question and relevant excerpts and sources from the documentation. Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.