Missing Dependency in C# Tensorflow.NET

[LouisSzeto]

Actual Behavior

C# Tensorflow.NET package only handles Tensorflow.Numpy.NDArray but not NumSharp.NDArray type. Lean is not able to handle Tensorflow.Numpy.NDArray with the following error:

System.TypeInitializationException: The type initializer for 'Tensorflow.Binding' threw an exception.
 ---> Tensorflow.RuntimeError: Tensorflow.NET cannot find a backend. Please install one of the following packages for your program: SciSharp.TensorFlow.Redist, SciSharp.TensorFlow.Redist-Linux-GPU, SciSharp.TensorFlow.Redist-Windows-GPU. For more details, please visit https://github.com/SciSharp/TensorFlow.NET. If it still not work after installing the backend, please submit an issue to https://github.com/SciSharp/TensorFlow.NET/issues
   at Tensorflow.tensorflow..ctor()
   at Tensorflow.Binding..cctor()
   --- End of inner exception stack trace ---
   at Tensorflow.Binding.get_tf()
   at Tensorflow.NumPy.AutoNumPyAttribute.OnEntry(MethodExecutionArgs args)
   at Tensorflow.NumPy.np.zeros(Shape shape, TF_DataType dtype)
   at Submission#6.<<Initialize>>d__0.MoveNext()
--- End of stack trace from previous location ---
   at Microsoft.CodeAnalysis.Scripting.ScriptExecutionState.RunSubmissionsAsync[TResult](ImmutableArray`1 precedingExecutors, Func`2 currentExecutor, StrongBox`1 exceptionHolderOpt, Func`2 catchExceptionOpt, CancellationToken cancellationToken)
   at Tensorflow.Binding.get_tf()
   at Tensorflow.NumPy.AutoNumPyAttribute.OnEntry(MethodExecutionArgs args)
   at Tensorflow.NumPy.np.zeros(Shape shape, TF_DataType dtype)
   at Submission#6.<<Initialize>>d__0.MoveNext()
--- End of stack trace from previous location ---
   at Microsoft.CodeAnalysis.Scripting.ScriptExecutionState.RunSubmissionsAsync[TResult](ImmutableArray`1 precedingExecutors, Func`2 currentExecutor, StrongBox`1 exceptionHolderOpt, Func`2 catchExceptionOpt, CancellationToken cancellationToken)

And also cannot load the Tensorflow backend

System.DllNotFoundException: Unable to load shared library 'tensorflow' or one of its dependencies. In order to help diagnose loading problems, consider setting the LD_DEBUG environment variable: libtensorflow: cannot open shared object file: No such file or directory
   at Tensorflow.c_api.TF_NewGraph()
   at Tensorflow.Graph..ctor()
   at Tensorflow.DefaultGraphStack.get_default()
   at Tensorflow.ops.get_default_graph()
   at Tensorflow.Keras.Utils.base_layer_utils.get_default_graph_uid_map()
   at Tensorflow.Keras.Utils.base_layer_utils.unique_layer_name(String name, Dictionary`2 name_uid_map, String[] avoid_names, Boolean zero_based)
   at Tensorflow.Keras.Engine.Layer._init_set_name(String name, Boolean zero_based)
   at Tensorflow.Keras.Engine.Layer.Initialize(LayerArgs args)
   at Tensorflow.Keras.Engine.Layer..ctor(LayerArgs args)
   at Tensorflow.Keras.Layers.InputLayer..ctor(InputLayerArgs args)
   at Tensorflow.Keras.Layers.LayersApi.Input(Shape shape, Int32 batch_size, String name, TF_DataType dtype, Boolean sparse, Tensor tensor, Boolean ragged, TypeSpec type_spec, Shape batch_input_shape, Shape batch_shape)
   at Tensorflow.Keras.KerasInterface.Input(Shape shape, Int32 batch_size, String name, TF_DataType dtype, Boolean sparse, Tensor tensor, Boolean ragged, TypeSpec type_spec, Shape batch_input_shape, Shape batch_shape)
   at Submission#7.<<Initialize>>d__0.MoveNext()
--- End of stack trace from previous location ---
   at Microsoft.CodeAnalysis.Scripting.ScriptExecutionState.RunSubmissionsAsync[TResult](ImmutableArray`1 precedingExecutors, Func`2 currentExecutor, StrongBox`1 exceptionHolderOpt, Func`2 catchExceptionOpt, CancellationToken cancellationToken)
   at Tensorflow.c_api.TF_NewGraph()
   at Tensorflow.Graph..ctor()
   at Tensorflow.DefaultGraphStack.get_default()
   at Tensorflow.ops.get_default_graph()
   at Tensorflow.Keras.Utils.base_layer_utils.get_default_graph_uid_map()
   at Tensorflow.Keras.Utils.base_layer_utils.unique_layer_name(String name, Dictionary`2 name_uid_map, String[] avoid_names, Boolean zero_based)
   at Tensorflow.Keras.Engine.Layer._init_set_name(String name, Boolean zero_based)
   at Tensorflow.Keras.Engine.Layer.Initialize(LayerArgs args)
   at Tensorflow.Keras.Engine.Layer..ctor(LayerArgs args)
...
   at Tensorflow.Keras.Layers.LayersApi.Input(Shape shape, Int32 batch_size, String name, TF_DataType dtype, Boolean sparse, Tensor tensor, Boolean ragged, TypeSpec type_spec, Shape batch_input_shape, Shape batch_shape)
   at Tensorflow.Keras.KerasInterface.Input(Shape shape, Int32 batch_size, String name, TF_DataType dtype, Boolean sparse, Tensor tensor, Boolean ragged, TypeSpec type_spec, Shape batch_input_shape, Shape batch_shape)
   at Submission#7.<<Initialize>>d__0.MoveNext()
--- End of stack trace from previous location ---
   at Microsoft.CodeAnalysis.Scripting.ScriptExecutionState.RunSubmissionsAsync[TResult](ImmutableArray`1 precedingExecutors, Func`2 currentExecutor, StrongBox`1 exceptionHolderOpt, Func`2 catchExceptionOpt, CancellationToken cancellationToken)

Potential Solution

Install the tensorflow backend (python?)

Reproducing the Problem

https://www.quantconnect.com/terminal/processCache?request=embedded_backtest_3b956d04556623f3e73a03e5b7905d58.html
Run the notebook

System Information

QC cloud

Checklist

• I have completely filled out this template
• I have confirmed that this issue exists on the current master branch
• I have confirmed that this is not a duplicate issue by searching issues

• I have provided detailed steps to reproduce the issue