Error in trainAutoencoder function

Hello,

when trying to train an autoencoder, no matter the input data, I get the following error:

Unable to resolve the name
Autoencoder.parseInputArguments.

Error in trainAutoencoder (line 107)
paramsStruct  =
Autoencoder.parseInputArguments(varargin{:});

I've tried running the example found in here, which should work just fine, but I get the aforementioned error, plus a line indicating

Error in filename (line 21)
autoenc1 = trainAutoencoder(xTrainImages,hiddenSize1,'MaxEpochs',400,'L2WeightRegularization',0.004,'SparsityRegularization',4,'SparsityProportion',0.15,'ScaleData',
false);

From what I understand, the error is inside the trainAutoencoder function, in the paramsStruct = Autoencoder.parseInputArguments(varargin{:}) line

if nargin > 1
    [varargin{:}] = convertStringsToChars(varargin{:});
end

paramsStruct  = Autoencoder.parseInputArguments(varargin{:});
autonet = Autoencoder.createNetwork(paramsStruct);
autoenc = Autoencoder.train(X, autonet, paramsStruct.UseGPU);
end

I've tried looking up the error online, but to no avail. If it's something obvious I just can't see what it is. I have the needed toolboxes installed, and I get this error in the 2019 and 2020 versions of MatLab.

If anyone could help me out, I would be extremely grateful

[Devineni Aslesha]

Hi

I could not replicate the issue when I tried running the 'Train Stacked Autoencoders for Image Classification' example mentioned in the question. Could you share the code used from the example that would help me to replicate the issue you are facing?

[Álvaro Pequeño Alonso]

Hi Devineni Aslesha thanks for answering

Here's my code, it's just the example mentioned:

clear all
close all

% Load the training data into memory
[xTrainImages,tTrain] = digitTrainCellArrayData;

% Display some of the training images
clf
for i = 1:20
    subplot(4,5,i);
    imshow(xTrainImages{i});
end

rng('default');
hiddenSize1 = 100;

autoenc1 = trainAutoencoder(xTrainImages,hiddenSize1,'MaxEpochs',400,'L2WeightRegularization',0.004,'SparsityRegularization',4,'SparsityProportion',0.15,'ScaleData', false);

view(autoenc1)

I'm not surprised you couldn't replicate it, I've tried asking my peers about it and nobody seems to be able to get the same error

[Devineni Aslesha]

I would suggest you to make sure to have the digitTrainCellArrayData.m file in the given path 'C:\Program Files\MATLAB\R2019a\toolbox\nnet\nndemos\nndatasets\digitTrainCellArrayData.m'. Also, try doing clear all before the code execution.

For more information, refer the following link.

https://www.mathworks.com/help/matlab/matlab_prog/ways-to-parse-function-inputs.html

[Álvaro Pequeño Alonso]

Thanks Devineni Aslesha, I got the idea that solved the problem from your comment

Answer 1

[Álvaro Pequeño Alonso]

Updating the post with the solution

Putting the path where the function .m file is at the top of the list, resolves the problem. I don't know why, but there it is

Home/Set Path, search for the path, and press Move to Top

To identify where the function is (the path) use "which 'function name' "

Answer 2

[Mohammad Eftekhari]

Sometimes this error is because of the name of the .m file in the running folder. it occurs when one of the .m files in the folder are the same as in the path direction.