Python Forum

Python 3.5
Tensorflow GPU

Hello, 
I am trying to perform a fft operation on a variable

x = tf.Variable([1,2,3,4,5],df.type =tf.float32)
y = tf.fft(x)
sess=tf.Session()
sess.run(model)
print(sess.run(d))

and it gives an error  

TypeError: Input 'input' of 'FFT' Op has type float32 that does not match expected type of complex64

So i tried converting the variable to complex number and it shows another error. 

a = tf.Variable ([1,2,3,4,5],dtype=tf.float32)
b= tf.Variable([0,0,0,0,0],dtype =tf.float32)
c = tf.complex(a,b)
d = tf.fft(c
model = init_all_variables()
sess = tf.Session()
sess.run(model)
print(sess.run(c))

InvalidArgumentError (see above for traceback): No OpKernel was registered to support Op 'FFT' with these attrs.  Registered kernels:
  <no registered kernels>

[[Node: FFT = FFT[](Complex)]]

Please help me to resolve the issue

Hello,
Your question/issue is specific to Tensorflow. So unless someone here is familiar with TensorFlow, then it's unlikely you will find an answer on this forum. Asking on TF Google Group was a better idea, so I hope guys there can help you out.
That said, I don't know how TF handles FFT input data, but FFT routines usually expect a "vector/array/list..." whose length is a power of two. Have you tried that out?
Good luck! JC