Model

Each sub-module of nextmv_sklearn has a convenience function for creating native sklearn model objects from nextmv.Options. This convenience function allows you to set up a model using the parameters that are customized through options. Notice that the return type is an sklearn.<Model>.

Consider the following examples, which make use of the classic diabetes dataset to fit a model and then make a prediction.

Dummy

Reference

Find the reference for the dummy.model module here.

import time

from nextmv_sklearn import dummy
from sklearn.datasets import load_diabetes

X, y = load_diabetes(return_X_y=True)

start_time = time.time()

options = dummy.DummyRegressorOptions().to_nextmv()
model = dummy.DummyRegressor(options)

fit = model.fit(X, y)
print(fit.predict(X[:1]))

Run the script:

$ python main.py --help
[152.13348416]

Ensemble

Reference

Find the reference for the ensemble.model module here.

• GradientBoostingRegressor

  import time
  
  from nextmv_sklearn import ensemble
  from sklearn.datasets import load_diabetes
  
  X, y = load_diabetes(return_X_y=True)
  
  start_time = time.time()
  
  options = ensemble.GradientBoostingRegressorOptions().to_nextmv()
  model = ensemble.GradientBoostingRegressor(options)
  
  fit = model.fit(X, y)
  print(fit.predict(X[:1]))
  

  Run the script:

  $ python main.py --help
  [200.87337372]
  

• RandomForestRegressor

  import time
  
  from nextmv_sklearn import ensemble
  from sklearn.datasets import load_diabetes
  
  X, y = load_diabetes(return_X_y=True)
  
  start_time = time.time()
  
  options = ensemble.RandomForestRegressorOptions().to_nextmv()
  model = ensemble.RandomForestRegressor(options)
  
  fit = model.fit(X, y)
  print(fit.predict(X[:1]))
  

  Run the script:

  $ python main.py --help
  [180.54]
  

Linear model

Reference

Find the reference for the linear_model.model module here.

import time

from nextmv_sklearn import linear_model
from sklearn.datasets import load_diabetes

X, y = load_diabetes(return_X_y=True)

start_time = time.time()

options = linear_model.LinearRegressionOptions().to_nextmv()
model = linear_model.LinearRegression(options)

fit = model.fit(X, y)
print(fit.predict(X[:1]))

Run the script:

$ python main.py --help
[206.11667725]

Neural network

Reference

Find the reference for the neural_network.model module here.

import time

from nextmv_sklearn import neural_network
from sklearn.datasets import load_diabetes

X, y = load_diabetes(return_X_y=True)

start_time = time.time()

options = neural_network.MLPRegressorOptions().to_nextmv()
model = neural_network.MLPRegressor(options)

fit = model.fit(X, y)
print(fit.predict(X[:1]))

Run the script:

$ python main.py -max_iter 2500
[195.53288524]

Tree

Reference

Find the reference for the tree.model module here.

import time

from nextmv_sklearn import tree
from sklearn.datasets import load_diabetes

X, y = load_diabetes(return_X_y=True)

start_time = time.time()

options = tree.DecisionTreeRegressorOptions().to_nextmv()
model = tree.DecisionTreeRegressor(options)

fit = model.fit(X, y)
print(fit.predict(X[:1]))

Run the script:

$ python main.py
[151.]