Persisting models with joblib or pickle

In this recipe, we're going to show how you can keep your model around for later use. For example, you might want to actually use a model to predict an outcome and automatically make a decision.

Getting ready

Create a dataset and train a classifier:

from sklearn.datasets import make_classification
from sklearn.tree import DecisionTreeClassifier

X, y = make_classification()
dt = DecisionTreeClassifier()
dt.fit(X, y)

How to do it...

1. Save the training work the classifier has done with joblib:

from sklearn.externals import joblib
joblib.dump(dt, "dtree.clf")

['dtree.clf']

Opening the saved model

2. Load the model with joblib. Make a prediction with a set of inputs:

from sklearn.externals import joblib
pulled_model = joblib.load("dtree.clf")
y_pred = pulled_model.predict(X)

We did not have to train the model again, and have saved a lot of training time. We simply reloaded it with joblib and made a prediction.

There's more...

You can also use the cPickle module in Python...