Using the myNeuralNetworkFunction generated from the Neural Network Toolbox App

Hi,

Forgive me if this is a simple problem but I can't seem to use the function generated after training a LSTM network in the app.

Once the app is finished it generates a script with the function, however, when i try to put a new data set into in it ([8xTS] 8 features over TS timesteps) the function rejects it with the error:

Cell contents reference from a non-cell array object.

Error in myNeuralNetworkFunction (line 75)

Xd1{ts} = mapminmax_apply(Xi{1,ts},x1_step1);

Here's the section of the script:

Xd1 = cell(1,3);

for ts=1:2

Xd1{ts} = mapminmax_apply(Xi{1,ts},x1_step1);

end

and the line I'm trying to input is

y = myNeuralNetworkFunction(input,2)

What am I doing wrong here?

Thanks in advance for your help

Answer 1

[Anshika Chaurasia]

Hi Damien,

• When the app is finished it generates a script with the function - myNeuralNetworkFunction(X,Xi,~) where Xi should be cell array.
• As you mentioned that you are using y = myNeuralNetworkFunction(input,2). Here second argument is not cell array i.e. Xi = 2.
• Hence provide second argument as cell array to myNeuralNetworkFunction.

%[Y,Xf,Af] = myNeuralNetworkFunction(X,Xi,~) takes these arguments:
%
%   X = 1xTS cell, 1 inputs over TS timesteps
%   Each X{1,ts} = 2xQ matrix, input #1 at timestep ts.
%
%   Xi = 1x2 cell 1, initial 2 input delay states.
%   Each Xi{1,ts} = 2xQ matrix, initial states for input #1.
%
%   Ai = 2x0 cell 2, initial 2 layer delay states.
%   Each Ai{1,ts} = 10xQ matrix, initial states for layer #1.
%   Each Ai{2,ts} = 2xQ matrix, initial states for layer #2.
%
% and returns:
%   Y = 1xTS cell of 1 outputs over TS timesteps.
%   Each Y{1,ts} = 2xQ matrix, output #1 at timestep ts.
%
%   Xf = 1x2 cell 1, final 2 input delay states.
%   Each Xf{1,ts} = 2xQ matrix, final states for input #1.
%
%   Af = 2x0 cell 2, final 0 layer delay states.
%   Each Af{1ts} = 10xQ matrix, final states for layer #1.
%   Each Af{2ts} = 2xQ matrix, final states for layer #2.
%
% where Q is number of samples (or series) and TS is the number of timesteps.

Refer to link for more understanding of error.