'PhysicsInformedNN' Error

[javadshamsicodes]

Hi, there is an error with 'PhysicsInformedNN' in model = PhysicsInformedNN(x0, u0, x1, u1, layers, dt, lb, ub, q, splitIdx1, splitIdx2, splitIdx3, Params).
Because 'PhysicInformed' object has no attribute 'initialize_NN'.

[ms140429]

Hi, there is an error with 'PhysicsInformedNN' in model = PhysicsInformedNN(x0, u0, x1, u1, layers, dt, lb, ub, q, splitIdx1, splitIdx2, splitIdx3, Params). Because 'PhysicInformed' object has no attribute 'initialize_NN'.

Hi javadshamsicodes,

Thanks for the feedback. However, the code runs well on my side. Can you please post your debug log or more details?

[javadshamsicodes]

Sorry for my late response.
I get the following debug when I run the otherParams.

AttributeError Traceback (most recent call last)
Input In [17], in <cell line: 1>()
41 splitIdx3 = 120
42 Params = otherParams()
---> 44 model = PhysicsInformedNN(x0, u0, x1, u1, layers, dt, lb, ub, q, splitIdx1, splitIdx2, splitIdx3, Params)
45 model.train(Params.AdamEpoch)
47 L = np.exp(model.sess.run(model.L))

Input In [3], in PhysicsInformedNN.init(self, x0, u0, x1, u1, layers, dt, lb, ub, q, splitIdx1, splitIdx2, splitIdx3, otherParams)
20 self.splitIdx3 = splitIdx3
21 # Initialize NN
---> 22 self.weights, self.biases = self.initialize_NN(layers)
23 self.otherParams = otherParams
25 self.L = tf.Variable([2], dtype=tf.float32)

AttributeError: 'PhysicsInformedNN' object has no attribute 'initialize_NN'

[ms140429]

Hi javadshamsicodes,

Thanks for the reply. The "initialized_NN" is defined on Line 100. Please check it. I guess you might run the code without initializing the class PhysicsInformedNN. Actually, you can directly run the code in Pycharm, which will automate everything.