Skip to main content
IBM Quantum Platform
이 페이지는 Qiskit SDK 의 이전 버전에서 가져온 것으로, 최신 버전에는 존재하지 않습니다. 최신 버전으로 업데이트하시기를 권장합니다. 자세한 정보는 릴리스 정보를 참조하십시오.

qiskit.pulse.library.GaussianSquareDrag

qiskit.pulse.library.GaussianSquareDrag(duration, amp, sigma, beta, width=None, angle=0.0, risefall_sigma_ratio=None, name=None, limit_amplitude=None)

GitHub

A square pulse with a Drag shaped rise and fall

This pulse shape is similar to GaussianSquare but uses Drag for its rise and fall instead of Gaussian. The addition of the DRAG component of the rise and fall is sometimes helpful in suppressing the spectral content of the pulse at frequencies near to, but slightly offset from, the fundamental frequency of the drive. When there is a spectator qubit close in frequency to the fundamental frequency, suppressing the drive at the spectator’s frequency can help avoid unwanted excitation of the spectator.

Exactly one of the risefall_sigma_ratio and width parameters has to be specified.

If risefall_sigma_ratio is not None and width is None:

risefall=risefall_sigma_ratio×sigmawidth=duration2×risefall\begin{aligned} \text{risefall} &= \text{risefall\_sigma\_ratio} \times \text{sigma}\\ \text{width} &= \text{duration} - 2 \times \text{risefall} \end{aligned}

If width is not None and risefall_sigma_ratio is None:

risefall=durationwidth2\text{risefall} = \frac{\text{duration} - \text{width}}{2}

Gaussian g(x,c,σ)g(x, c, σ) and lifted gaussian g(x,c,σ)g'(x, c, σ) curves can be written as:

g(x,c,σ)=exp(12(xc)2σ2)g(x,c,σ)=g(x,c,σ)g(1,c,σ)1g(1,c,σ)\begin{aligned} g(x, c, σ) &= \exp\Bigl(-\frac12 \frac{(x - c)^2}{σ^2}\Bigr)\\ g'(x, c, σ) &= \frac{g(x, c, σ)-g(-1, c, σ)}{1-g(-1, c, σ)} \end{aligned}

From these, the lifted DRAG curve d(x,c,σ,β)d'(x, c, σ, β) can be written as

d(x,c,σ,β)=g(x,c,σ)×(1+1j×β×(xcσ2))d'(x, c, σ, β) = g'(x, c, σ) \times \Bigl(1 + 1j \times β \times \Bigl(-\frac{x - c}{σ^2}\Bigr)\Bigr)

The lifted gaussian square drag pulse f(x)f'(x) is defined as:

f(x)={A×d(x,risefall,sigma,beta)x<risefallArisefallx<risefall+widthA××d(x(risefall+width),risefall,sigma,beta)risefall+widthx\begin{aligned} f'(x) &= \begin{cases} \text{A} \times d'(x, \text{risefall}, \text{sigma}, \text{beta}) & x < \text{risefall}\\ \text{A} & \text{risefall} \le x < \text{risefall} + \text{width}\\ \text{A} \times \times d'( x - (\text{risefall} + \text{width}), \text{risefall}, \text{sigma}, \text{beta} ) & \text{risefall} + \text{width} \le x \end{cases}\\ \end{aligned}

where A=amp×exp(i×angle)\text{A} = \text{amp} \times \exp\left(i\times\text{angle}\right).

Deprecated since version 1.3

The function qiskit.pulse.library.symbolic_pulses.GaussianSquareDrag() is deprecated as of Qiskit 1.3. It will be removed in Qiskit 2.0. The entire Qiskit Pulse package is being deprecated and will be moved to the Qiskit Dynamics repository: https://github.com/qiskit-community/qiskit-dynamics

Parameters

  • duration (int |ParameterExpression) – Pulse length in terms of the sampling period dt.
  • amp (float |ParameterExpression) – The amplitude of the DRAG rise and fall and of the square pulse.
  • sigma (float |ParameterExpression) – A measure of how wide or narrow the DRAG risefall is; see the class docstring for more details.
  • beta (float |ParameterExpression) – The DRAG correction amplitude.
  • width (float |ParameterExpression | None) – The duration of the embedded square pulse.
  • angle (float |ParameterExpression | None) – The angle in radians of the complex phase factor uniformly scaling the pulse. Default value 0.
  • risefall_sigma_ratio (float |ParameterExpression | None) – The ratio of each risefall duration to sigma.
  • name (str | None) – Display name for this pulse envelope.
  • limit_amplitude (bool | None) – If True, then limit the amplitude of the waveform to 1. The default is True and the amplitude is constrained to 1.

Returns

ScalableSymbolicPulse instance.

Raises

PulseError – When width and risefall_sigma_ratio are both empty or both non-empty.

Return type

ScalableSymbolicPulse

이 페이지가 도움이 되었습니까?
GitHub에서 버그, 오타를 보고하거나 컨텐츠를 요청하십시오.