Get upper bound for the variable corresponding to layer, node If b=True, b variable Otherwise, f variable Args:

Function to find underlying constants or variables representing operation Args: op: (tf.op) operation to find values for

Get variable number corresponding to layer, node If b=True, b variable Otherwise, f variable If layer is 0 (inpu

Serializes the inputQuery in the given filename Args: filename: (string) file to write serialized inputQuery

Create a random input for the network. The value for each input variable is chosen uniformly at random between the lower and the upper bounds

Load network and evaluate testInputs with and without Marabou Args: filename (str): name of network file without path

Get lower bound for the variable corresponding to layer, node If b=True, b variable Otherwise, f variable Args:

Get the score of the PLConstraint */

* Return the current (top) scope */

Function to check whether lower bound for a variable is known Args: x (int): Variable to check

Parses the provided ONNX graph into constraints which are stored in the query argument. Args: query: the query to which

Function representing transpose Args: node (node): ONNX node representing transpose operation :meta private:

Function to check whether upper bound for a variable is known Args: x (int): Variable to check

Reset values to represent empty network

Create an options object for how Marabou should solve the query Args: numWorkers (int, optional): Number of workers to use in Split-and-C