Graph.gaussian_convolution_kernel(std, offset=0)

Creates a convolution kernel representing a normalized Gaussian.

Use this kernel to allow frequencies in the range roughly determined by its width, and progressively suppress components outside that range.

  • std (float or Tensor) – Standard deviation \(\sigma\) of the Gaussian in the time domain. The standard deviation in the frequency domain is its inverse, so that a high value of this parameter lets fewer frequencies pass.

  • offset (float or Tensor, optional) – Center \(\mu\) of the Gaussian distribution in the time domain. Use this to offset the signal in time. Defaults to 0.


A node representing a Gaussian function to use in a convolution.

Return type



The Gaussian that this node represents is normalized in the time domain:


In the frequency domain, this Gaussian has standard deviation \(\omega_c= \sigma^{-1}\). The filter it represents therefore passes frequencies roughly in the range \([-\omega_c, \omega_c]\).

See also


Create an Stf by convolving a Pwc with a kernel.


Create a convolution kernel representing the sinc function.


Filter a signal by convolving it with a Gaussian kernel.

>>> gaussian_kernel = graph.gaussian_convolution_kernel(std=1.0, offset=3.0)
>>> gaussian_kernel
<ConvolutionKernel: operation_name="gaussian_convolution_kernel">
>>> signal
<Pwc: name="alpha", operation_name="pwc_signal", value_shape=(), batch_shape=()>
>>> filtered_signal = graph.convolve_pwc(pwc=signal, kernel=gaussian_kernel)
>>> filtered_signal
<Stf: operation_name="convolve_pwc", value_shape=(), batch_shape=()>

Refer to the How to characterize the bandwidth of a transmission line using a qubit as a probe user guide to find the example in context.