Skip to content

Saving and Loading a model

Saving or Loading a model is very straighforward in pySurvival.

Saving

To save a model, use the function save_model and provide the full path of the future location of the file as the argument; the model is then compressed into a .zip file. The function is located at pysurvival.utils.save_model.

API

save_model - Save and compress the model and its parameters into a .zip file

save_model(path_file)

Parameters:

  • model : Pysurvival object -- Pysurvival model

  • path_file, str -- full address of the file where the model will be saved

Example

# Importing modules
from pysurvival.models.svm import KernelSVMModel
from pysurvival.datasets import Dataset

# Loading and splitting a simple example into train/test sets
X_train, T_train, E_train, \
    X_test, T_test, E_test = Dataset('simple_example').load_train_test()

# Let's assume we want to build the following SVM model
svm_model = KernelSVMModel('gaussian')
svm_model.fit(X_train, T_train, E_train, init_method='glorot_uniform',
              l2_reg = 1e-5, lr = 0.5)

# Let's now save our model
from pysurvival.utils import save_model
save_model(svm_model, '/Users/xxx/Desktop/svm_model_2018_08_26.zip')

Loading

To load a model, use the function load_model and provide the full path of the location of the file as the argument. The function is located at pysurvival.utils.load_model.

API

load_model - Load the model and its parameters from a .zip file

load_model(path_file)

Parameters:

  • path_file : str -- full address of the file where the model will be loaded from

Example

# Let's assume we have built and saved a SVM model at the following location
# /Users/xxx/Desktop/svm_model_2018_08_26.zip

from pysurvival.utils import load_model
svm_model = load_model('/Users/xxx/Desktop/svm_model_2018_08_26.zip')