Var¶

classonnx_array_api.npx.npx_var.Var(*inputs:List[Any],op:Callable|str|Tuple[str,str]|FunctionProto|ModelProto|NodeProto|None=None,dtype:type|DType|None=None,inline:bool=False,n_var_outputs:int=1,input_indices:List[int]|None=None,**kwargs)[source]¶

Defines a variable, a result…

Parameters:

• inputs – list of inputs

• op – apply on operator on the inputs

• inline – True to reduce the use of function and inlinesmall functions, this only applies ifop is a function

• n_var_outputs – number of the operator outputs

• input_indices – to select a specific output from the inputoperator

• kwargs – operator attributes

Private attribute:

Parameters:

onnx_input_type – names given to the variables

propertyT:Var¶

Transpose.

propertyannotation¶

Returns a type if known for the Var itself.

astype(dtype)→Var[source]¶

Cast

copy()→Var[source]¶

Returns a copy of self (use of Identity node).

flatten()→Var[source]¶

Flattens a matrix (seenumpy.ndarray.flatten()).

Parameters:

axis – only flatten from axis to the end.

Returns:

Var

get(index:int)→Var[source]¶

If an operator or a function returns more than one output,this takes only one.

Parameters:

index – index of the output to select

Returns:

Var

propertyis_function¶

Tells if this variable encapsulate a function.

max(axis:TensorType_I__DT7=None,keepdims:ParTypeint=0)→Var[source]¶

Seenumpy.max().

mean(axis:OptParTypeTupleType_int=None,keepdims:ParTypeint=0)→Var[source]¶

Seenumpy.mean().

min(axis:TensorType_I__DT7=None,keepdims:ParTypeint=0)→Var[source]¶

Seenumpy.min().

prod(axis:TensorType_I__DT7=None,keepdims:ParTypeint=0)→Var[source]¶

Seenumpy.prod().

reduce_function(reduce_op,axis:OptTensorType_I__DT7=None,keepdims:ParTypeint=0)→Var[source]¶

Seenumpy.sum() or any other reduce function.

replace_inputs(new_inputs:List[Var],input_indices:List[int]|None=None)→Var[source]¶

Replaces inputs by new ones. It creates a copy.It is needed when inlining functions.

reshape(shape:Var)→Var[source]¶

Reshape

propertyself_var¶

Returns itself or the variable corresponding to itsstate after a call to__setitem__.

set_onnx_name(prefix:str)[source]¶

Forces this variable to get this name during

Parameters:

prefix – prefix

propertyshape:Var¶

Shape

sum(axis:TensorType_I__DT7=None,keepdims:ParTypeint=0)→Var[source]¶

Seenumpy.sum().

to_onnx(target_opsets:Dict[str,int]|None=None,as_function:bool=False,name:str|None=None,domain:str|None=None,attributes:List[str]|None=None,constraints:Dict[Any,TensorType]|None=None,ir_version:int|None=None)→ModelProto|FunctionProto|List[Any][source]¶

Converts the recursive graph to ONNX.

Parameters:

• target_opsets – dictionary{opset: version}

• as_function – conversion toonnx.FunctionProtooronnx.ModelProto

• name – function name ifas_function is True

• domain – function domain ifas_function is True

• attributes – function attributes if any

• constraints – specifies a precise type for the typeconstraints when a function allows more than one type,this works if there is only one variable to be converted

Returns:

ModelProto, FunctionProto

Cst, Input¶

classonnx_array_api.npx.npx_var.Cst(cst:Any)[source]¶

Defines a constant.

classonnx_array_api.npx.npx_var.Input(name:str|None=None,annotation:type|None=None)[source]¶

Defines an input, a placeholder.

Parameters:

• name – input name or None if undefined

• annotation – annotation if any is available

propertyannotation¶

Returns a type if known for the Var itself.

ManyIdentity¶

classonnx_array_api.npx.npx_var.ManyIdentity(*inputs,input_indices=None)[source]¶

Holds several instances ofVar.

to_onnx(target_opsets:Dict[str,int]|None=None,as_function:bool=False,name:str|None=None,domain:str|None=None,attributes:List[str]|None=None,constraints:Dict[Any,TensorType]|None=None,ir_version:int|None=None)→ModelProto|FunctionProto|List[Any][source]¶

Converts the recursive graph to ONNX.

Parameters:

• target_opsets – dictionary{opset: version}, if None,it is replaced byDEFAULT_OPSETS

• as_function – conversion toonnx.FunctionProtooronnx.ModelProto

• name – function name ifas_function is True

• domain – function domain ifas_function is True

• attributes – function attributes if any

• constraints – specifies a precise type for the typeconstraints when a function allows more than one type,this works if there is only one variable to be converted

Returns:

ModelProto, FunctionProto

Par¶

classonnx_array_api.npx.npx_var.Par(name:str,dtype:ParType,value:Any|None=None,parent_op:Tuple[str,str,int]|None=None)[source]¶

Defines a named parameter.

Parameters:

• name – parameter name

• dtype – parameter type (bool, int, str, float)

• value – value of the parameter if known

• parent_op – node type it belongs to

propertyonnx_type¶

Returns the corresponding onnx type.