Package com.helix.core.ml

Class OnnxFeatureExtractor

java.lang.Object

com.helix.core.ml.OnnxFeatureExtractor

public class OnnxFeatureExtractor
extends Object

Extracts and transforms runtime execution context features into native ONNX tensor buffers.

Validates schema alignment against OnnxModelDescriptor.inputFeatures(), safely coerces mixed numeric and boolean types into single-precision float representations, and constructs zero-copy 2D arrays ready for OnnxTensor allocation.

Constructor Summary

Constructors

Constructor	Description
OnnxFeatureExtractor()

Method Summary

Modifier and Type	Method	Description
ai.onnxruntime.OnnxTensor	createTensor(ai.onnxruntime.OrtEnvironment env, OnnxModelDescriptor descriptor, ExecutionContext context)	Constructs a native OnnxTensor for a single-row inference request.
ai.onnxruntime.OnnxTensor	createTensor(ai.onnxruntime.OrtEnvironment env, OnnxModelDescriptor descriptor, List<ExecutionContext> contexts)	Constructs a native OnnxTensor for a batch inference request.
float[][]	extractBatchFeatures(OnnxModelDescriptor descriptor, List<ExecutionContext> contexts)	Extracts an ordered 2D batch feature matrix for multiple execution contexts.
float[]	extractFeatures(OnnxModelDescriptor descriptor, ExecutionContext context)	Extracts an ordered feature vector for a single execution context.

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Constructor Details

OnnxFeatureExtractor

public OnnxFeatureExtractor()

Method Details

extractFeatures

public float[] extractFeatures(OnnxModelDescriptor descriptor,
 ExecutionContext context)

Extracts an ordered feature vector for a single execution context.

Parameters:

descriptor - target ONNX model descriptor

context - runtime evaluation context

Returns:

1D float array of model input features

Throws:

IllegalArgumentException - if a required feature is missing or cannot be coerced to float

extractBatchFeatures

public float[][] extractBatchFeatures(OnnxModelDescriptor descriptor,
 List<ExecutionContext> contexts)

Extracts an ordered 2D batch feature matrix for multiple execution contexts.

Parameters:

descriptor - target ONNX model descriptor

contexts - list of execution contexts

Returns:

2D float array with shape [batch_size, feature_count]

createTensor

public ai.onnxruntime.OnnxTensor createTensor(ai.onnxruntime.OrtEnvironment env,
 OnnxModelDescriptor descriptor,
 ExecutionContext context)
                                       throws ai.onnxruntime.OrtException

Constructs a native OnnxTensor for a single-row inference request.

Parameters:

env - active ONNX Runtime environment

descriptor - target ONNX model descriptor

context - single evaluation context

Returns:

allocated native ONNX tensor buffer with shape [1, feature_count]

Throws:

ai.onnxruntime.OrtException - if native tensor allocation fails

createTensor

public ai.onnxruntime.OnnxTensor createTensor(ai.onnxruntime.OrtEnvironment env,
 OnnxModelDescriptor descriptor,
 List<ExecutionContext> contexts)
                                       throws ai.onnxruntime.OrtException

Constructs a native OnnxTensor for a batch inference request.

Parameters:

env - active ONNX Runtime environment

descriptor - target ONNX model descriptor

contexts - batch of evaluation contexts

Returns:

allocated native ONNX tensor buffer with shape [batch_size, feature_count]

Throws:

ai.onnxruntime.OrtException - if native tensor allocation fails