isolation-forest-onnx

A converter for the LinkedIn Spark/Scala isolation forest model format to ONNX format for broad portability across platforms and languages.

Note: ONNX conversion is supported for the standard IsolationForestModel only. The ExtendedIsolationForestModel uses hyperplane-based splits that are not compatible with the axis-aligned tree ensemble representation used by the ONNX converter.

Installation

pip install isolation-forest-onnx

It is recommended to use the same version of the converter as the version of the isolation-forest library used to train the model.

Converting a trained model to ONNX

import os
from isolationforestonnx.isolation_forest_converter import IsolationForestConverter

# Path where the trained IsolationForestModel was saved in Scala
path = '/user/testuser/isolationForestWriteTest'

# Get model data path
data_dir_path = path + '/data'
avro_model_file = os.listdir(data_dir_path)
model_file_path = data_dir_path + '/' + avro_model_file[0]

# Get model metadata file path
metadata_dir_path = path + '/metadata'
metadata_file = os.listdir(metadata_dir_path)
metadata_file_path = metadata_dir_path + '/' + metadata_file[0]

# Convert the model to ONNX format (returns the ONNX model in memory)
converter = IsolationForestConverter(model_file_path, metadata_file_path)
onnx_model = converter.convert()

# Convert and save the model in ONNX format
onnx_model_path = '/user/testuser/isolationForestWriteTest.onnx'
converter.convert_and_save(onnx_model_path)

Using the ONNX model for inference

import numpy as np
import onnx
from onnxruntime import InferenceSession

onnx_model_path = '/user/testuser/isolationForestWriteTest.onnx'
dataset_path = 'shuttle.csv'

# Load data
input_data = np.loadtxt(dataset_path, delimiter=',')
num_features = input_data.shape[1] - 1
last_col_index = num_features

# The last column is the label column
input_dict = {'features': np.delete(input_data, last_col_index, 1).astype(dtype=np.float32)}

# Load the ONNX model and run inference
onx = onnx.load(onnx_model_path)
sess = InferenceSession(onx.SerializeToString())
res = sess.run(None, input_dict)

# Print scores
outlier_scores = res[0]
print(np.transpose(outlier_scores[:10])[0])

License

BSD 2-Clause License. See LICENSE for details.

Metadata

Release files for isolation-forest-onnx 4.1.8

For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.

Source distribution (sdist)

Source distribution for isolation-forest-onnx 4.1.8
File	Size	Uploaded
isolation_forest_onnx-4.1.8.tar.gz	9.8 kB	Aug 4, 2026	Details

Built distribution (wheel)

Table of built distributions (wheels) for isolation-forest-onnx 4.1.8
File	Interpreter	ABI	Platform	Reset
isolation_forest_onnx-4.1.8-py3-none-any.whl 6.5 kB Aug 4, 2026	Python 3	none	any	Details

Total release size: 16.2 kB

Release files / isolation_forest_onnx-4.1.8.tar.gz

Download URL	isolation_forest_onnx-4.1.8.tar.gz
Size	9.8 kB
Tags	Source
SHA-256 checksum How to use checksums	`263eb83d9a260050eb982cd125a9ffbc586088c3ed994e218567d7078ce22609`
BLAKE2b-256 checksum How to use checksums	`2d6bd39ae057a112197d127fabfcf765c9a7567b7b94136a14e6bb55f6941594`
Upload date	Aug 4, 2026
Uploaded using Trusted Publishing? What is trusted publishing?	No
Uploaded via	`twine/6.2.0 CPython/3.12.13`

Release files / isolation_forest_onnx-4.1.8-py3-none-any.whl

Download URL	isolation_forest_onnx-4.1.8-py3-none-any.whl
Size	6.5 kB
Tags	Python 3
SHA-256 checksum How to use checksums	`0e111402a63b1af2c1f562fe30bcef175df26cdf5a8bde624dbcfc7403d1eb11`
BLAKE2b-256 checksum How to use checksums	`846d94a7b6ef7f7363cb5c28d343152115e6491265a1eb277c2a5e7cb3fac41b`
Upload date	Aug 4, 2026
Uploaded using Trusted Publishing? What is trusted publishing?	No
Uploaded via	`twine/6.2.0 CPython/3.12.13`