Can somebody please guide me how to remove following error

Error in sequenceInputLayer (line 26)

inputArguments = iParseInputArguments(varargin{:});

Error in lstm_2 (line 62)

sequenceInputLayer(inputsize,'Name','ip')

[Walter Roberson]

we need the full error message

[ALI]

[Jan]

We still do not see the complete message. There should be some lines on top of the message telling, what the problem is. You show only the part, which mentions, where the error occurs.

Prefer to copy&paste the text instead of posting huge screen shots.

[ALI]

this is the code where it shows error

%define LSTM
inputsize=[1,rt,ct];
NOF=6;
NOR=1;
NHU=200;

layers =[
    sequenceInputLayer(inputsize,'Name','ip')
    %       sequenceinputlayer(NOF)
    %       flattenlayer([rt 3 2],rt*3*2)
    flattenlayer('Name','F1')
    lstmLayer(NHU,'OutputMode','sequence','Name','LSTM')      %last for cell data
    fullyConnectedLayer(NOR, 'Name','fC')
    regressionLayer('Name','RL')];
lgraph=layerGraph(layers);
%   lgraph=connectLayers(lgraph,'relu_1','add/in2');
plot(lgraph);

[ALI]

this is the error ]

% Parse the input arguments.
inputArguments = iParseInputArguments(varargin{:});

% Create an internal representation of a sequence input layer.
internalLayer = nnet.internal.cnn.layer.SequenceInput( ...
    inputArguments.Name, ...
    inputArguments.InputSize );

Answer 1

[Harsha]

MATLAB R2018a and R2018b only support InputSize argument in sequenceInputLayer as scalar.

InputSize as vector is supported from MATLAB R2019a release.

Refer to this documentation links regarding the same

R2018b - https://in.mathworks.com/help/releases/R2018b/deeplearning/ref/nnet.cnn.layer.sequenceinputlayer.html

R2019a - https://in.mathworks.com/help/releases/R2019a/deeplearning/ref/nnet.cnn.layer.sequenceinputlayer.html