MATLAB Answers

Getting the error: Undefined function 'network' for input arguments of type 'double' while generating neural network using newgrnn

20 views (last 30 days)
Anjana Krishnan
Anjana Krishnan on 6 Dec 2016
Line of code is : nn_t = newgrnn(ip,op,0.03);
Tried: which newgrnn C:\Program Files\MATLAB\R2016b\toolbox\nnet\nnet\nnnetwork\newgrnn.m >> which network C:\Program Files\MATLAB\R2016b\toolbox\nnet\nnet\@network\network.m % network constructor, Has no license available

  0 Comments

Sign in to comment.

Answers (1)

Sid Jhaveri
Sid Jhaveri on 8 Dec 2016
Hi Anjana,
This looks like licensing issue. Please try out the following steps:
1) Execute the command given below:
>>license('checkout','Neural_Network_Toolbox')
Depending upon the output of the above command, follow the next steps.
2) If output is 1, then this mean the toolbox with a valid license has been installed properly. In this case, try the following commands:
>>rehash toolboxcache
>>restoredefaultpath
If you are still facing the same issue, contact MathWorks Installation Support.
3) The output throws an error message and the error number is "39". This means that you have a network license and the administrator has not given you the permission to work with this toolbox. I would suggest asking the administrator of your organization to grant you the permission for the same.
4) In case of any other scenario, contact MathWorks Installation Support.

  2 Comments

Santhosh Tamilarasan
Santhosh Tamilarasan on 7 Aug 2019
Hi, I am having the same issue too!
I tried all the options given by Sid in the first answer, and still getting the same error!
How did you fix the path mismatch problem? I am getting the same error both in 2016a, 2016b. Even the democode demogrn1.m throws the error.
>> demogrn1
Undefined function 'network' for input arguments of type 'double'.
Error in newgrnn>create_network (line 109)
net = network(1,2,[1;0],[1;0],[0 0;1 0],[0 1]);
Error in newgrnn (line 66)
net = create_network(param);
Error in demogrn1 (line 27)
net = newgrnn(X,T,spread);

Sign in to comment.

Sign in to answer this question.