👉 Use MLFlow API for Scoring¶

Dataset For Scoring¶

In [1]:

from pycaret.datasets import get_data
data = get_data('insurance', verbose=False)
data.drop('charges', axis=1, inplace=True)
data.head()

Out[1]:

	age	sex	bmi	children	smoker	region
0	19	female	27.900	0	yes	southwest
1	18	male	33.770	1	no	southeast
2	28	male	33.000	3	no	southeast
3	33	male	22.705	0	no	northwest
4	32	male	28.880	0	no	northwest

Define MLFlow Scoring Function¶

In [2]:

import mlflow

def score_model(data, model_name, model_version):
    
    mlflow.set_tracking_uri("sqlite:///mlruns.db")
    model_uri = "models:/{}/{}".format(model_name, model_version)
    model = mlflow.pyfunc.load_model(model_uri)
    return model.predict(data)

In [3]:

score_model(data, 'my_first_model', 1)

Out[3]:

array([18894.26007319,  3698.2875344 ,  6029.27157845, ...,
        2442.28835297,  2613.21866387, 28782.04091164])

In [4]:

score_model(data, 'my_first_model', 2)

Out[4]:

array([16831.03017578,  2908.97255127,  4915.3686792 , ...,
        2035.52375305,  2061.27198486, 28688.36953125])

In [5]:

score_model(data, 'my_first_model', 3)

Out[5]:

array([25119.76880225,  4104.82323122,  6043.30438595, ...,
        4195.51717011,  1111.38184836, 36858.02738907])

👉 Alteratively ...¶

In [6]:

from pycaret.regression import load_model, predict_model
l = load_model('mlruns/1/c391ad05516442eb85c5defd46c27931/artifacts/model/model')

Transformation Pipeline and Model Successfully Loaded

In [7]:

print(l)

Pipeline(steps=[('dtypes',
                 DataTypes_Auto_infer(display_types=False,
                                      ml_usecase='regression',
                                      target='charges')),
                ('imputer',
                 Simple_Imputer(categorical_strategy='not_available',
                                fill_value_categorical=None,
                                fill_value_numerical=None,
                                numeric_strategy='mean',
                                target_variable=None)),
                ('new_levels1',
                 New_Catagorical_Levels_in_TestData(replacement_strategy='leas...
                ('binn', 'passthrough'), ('rem_outliers', 'passthrough'),
                ('cluster_all', 'passthrough'),
                ('dummy', Dummify(target='charges')),
                ('fix_perfect', Remove_100(target='charges')),
                ('clean_names', Clean_Colum_Names()),
                ('feature_select', 'passthrough'), ('fix_multi', 'passthrough'),
                ('dfs', 'passthrough'), ('pca', 'passthrough'),
                ['trained_model', GradientBoostingRegressor(random_state=123)]])

In [8]:

predictions = predict_model(l, data=data)
predictions.head()

Out[8]:

	age	sex	bmi	children	smoker	region	Label
0	19	female	27.900	0	yes	southwest	18894.260073
1	18	male	33.770	1	no	southeast	3698.287534
2	28	male	33.000	3	no	southeast	6029.271578
3	33	male	22.705	0	no	northwest	8958.189116
4	32	male	28.880	0	no	northwest	3900.039002

In [ ]: