def prepare(self):
# Create files generated by exporter in the temp dir.
code = export_to_java(self.model, class_name=self.class_name)
code_file_name = self._resource_tmp_dir / f"{self.class_name}.java"
utils.write_content_to_file(code, code_file_name)
# Move Executor.java to the same temp dir.
module_path = Path(__file__).absolute().parent
executor_path = self._resource_tmp_dir / "Executor.java"
copyfile(module_path / "Executor.java", executor_path)
# Compile all files together.
subprocess.call(
[str(self._javac_bin),
str(code_file_name),
str(executor_path)])
def prepare(self):
# Create files generated by exporter in the temp dir.
code = export_to_java(self.model, class_name=self.class_name)
code_file_name = os.path.join(self._resource_tmp_dir, f"{self.class_name}.java")
with open(code_file_name, "w") as f:
f.write(code)
# Move Executor.java to the same temp dir.
module_path = os.path.dirname(__file__)
shutil.copy(os.path.join(module_path, "Executor.java"), self._resource_tmp_dir)
# Compile all files together.
subprocess.call([
self._javac_bin,
code_file_name,
os.path.join(self._resource_tmp_dir, "Executor.java")
])
import os
import m2cgen as m2c
from sklearn.datasets import load_diabetes
from sklearn.tree import DecisionTreeRegressor
# Load data
X, y = load_diabetes(return_X_y=True)
# Create + Train ML model
tree = DecisionTreeRegressor()
tree.fit(X, y)
# Translate to .java code
code = m2c.export_to_java(tree,
package_name=None,
class_name="DecisionTreeModel",
indent=4,
function_name="score")
# Save .java code
out_path = os.path.join('test/model_small', 'DecisionTreeModel.java')
with open(out_path, 'w') as f:
f.write(code)
# https://github.com/BayesWitnesses/m2cgen import m2cgen as m2c code = m2c.export_to_java(model)
