Quantum Fourier Transform (QFT)

The quantum Fourier transform is a core component of many quantum algorithms — such as phase estimation and Shor’s factoring — provided ready-made in the algorithm library.

Overview

QFT operators overview

Operator

Operation

Unitarity class

QFT

Quantum Fourier transform

BaseOperator

InverseQFT

Inverse quantum Fourier transform

BaseOperator

Mathematical Definition

The quantum Fourier transform is defined as:

\[QFT|x\rangle = \frac{1}{\sqrt{2^n}} \sum_{y=0}^{2^n-1} e^{2\pi i xy / 2^n} |y\rangle\]

The inverse quantum Fourier transform:

\[QFT^\dagger|y\rangle = \frac{1}{\sqrt{2^n}} \sum_{x=0}^{2^n-1} e^{-2\pi i xy / 2^n} |x\rangle\]

—

QFT

class pysparq.QFT

Quantum Fourier Transform on a register.

Applies the QFT to transform between computational and Fourier bases. Commonly used in phase estimation and Shor’s algorithm.

Parameters:

reg_name – Name of the register to transform (str) or register ID (int).

Example

QFT(“data”)(state) # Apply QFT # … computation … InverseQFT(“data”)(state) # Apply inverse QFT

clear_control_all_ones(self: pysparq._core.QFT) → None

Clear all control conditions of the specified type.

clear_control_by_bit(self: pysparq._core.QFT) → None

Clear all control conditions of the specified type.

clear_control_by_value(self: pysparq._core.QFT) → None

Clear all control conditions of the specified type.

clear_control_nonzeros(self: pysparq._core.QFT) → None

Clear all control conditions of the specified type.

property condition_variable_all_ones
property condition_variable_by_bit
property condition_variable_by_value
property condition_variable_nonzeros
conditioned_by_all_ones(*args, **kwargs)

Overloaded function.

  1. conditioned_by_all_ones(self: pysparq._core.QFT, cond: str) -> pysparq._core.QFT

Condition this operation on registers where all bits are 1.

Calling this method replaces prior all-ones conditions. Pass the list overload to require several registers simultaneously; conditions of different kinds are combined with logical AND.

Parameters:
  • cond – Register name (str) or ID (int) to condition on.

  • conds – List of register names or IDs for multi-condition.

Returns:

Self, for method chaining.

  1. conditioned_by_all_ones(self: pysparq._core.QFT, conds: collections.abc.Sequence[str]) -> pysparq._core.QFT

Condition this operation on registers where all bits are 1.

Calling this method replaces prior all-ones conditions. Pass the list overload to require several registers simultaneously; conditions of different kinds are combined with logical AND.

Parameters:
  • cond – Register name (str) or ID (int) to condition on.

  • conds – List of register names or IDs for multi-condition.

Returns:

Self, for method chaining.

  1. conditioned_by_all_ones(self: pysparq._core.QFT, cond: typing.SupportsInt | typing.SupportsIndex) -> pysparq._core.QFT

Condition this operation on registers where all bits are 1.

Calling this method replaces prior all-ones conditions. Pass the list overload to require several registers simultaneously; conditions of different kinds are combined with logical AND.

Parameters:
  • cond – Register name (str) or ID (int) to condition on.

  • conds – List of register names or IDs for multi-condition.

Returns:

Self, for method chaining.

  1. conditioned_by_all_ones(self: pysparq._core.QFT, conds: collections.abc.Sequence[typing.SupportsInt | typing.SupportsIndex]) -> pysparq._core.QFT

Condition this operation on registers where all bits are 1.

Calling this method replaces prior all-ones conditions. Pass the list overload to require several registers simultaneously; conditions of different kinds are combined with logical AND.

Parameters:
  • cond – Register name (str) or ID (int) to condition on.

  • conds – List of register names or IDs for multi-condition.

Returns:

Self, for method chaining.

conditioned_by_bit(*args, **kwargs)

Overloaded function.

  1. conditioned_by_bit(self: pysparq._core.QFT, cond: str, pos: typing.SupportsInt | typing.SupportsIndex) -> pysparq._core.QFT

Condition this operation on a specific bit position.

Calling this method replaces prior bit conditions. Pass the list-of-pairs overload to require several bits simultaneously; conditions of different kinds are combined with logical AND.

Parameters:
  • cond – Register name (str) or ID (int).

  • pos – Bit position to check (0-indexed).

  • conds – List of (register, position) pairs.

Returns:

Self, for method chaining.

  1. conditioned_by_bit(self: pysparq._core.QFT, conds: collections.abc.Sequence[tuple[str, typing.SupportsInt | typing.SupportsIndex]]) -> pysparq._core.QFT

Condition this operation on a specific bit position.

Calling this method replaces prior bit conditions. Pass the list-of-pairs overload to require several bits simultaneously; conditions of different kinds are combined with logical AND.

Parameters:
  • cond – Register name (str) or ID (int).

  • pos – Bit position to check (0-indexed).

  • conds – List of (register, position) pairs.

Returns:

Self, for method chaining.

  1. conditioned_by_bit(self: pysparq._core.QFT, cond: typing.SupportsInt | typing.SupportsIndex, pos: typing.SupportsInt | typing.SupportsIndex) -> pysparq._core.QFT

Condition this operation on a specific bit position.

Calling this method replaces prior bit conditions. Pass the list-of-pairs overload to require several bits simultaneously; conditions of different kinds are combined with logical AND.

Parameters:
  • cond – Register name (str) or ID (int).

  • pos – Bit position to check (0-indexed).

  • conds – List of (register, position) pairs.

Returns:

Self, for method chaining.

  1. conditioned_by_bit(self: pysparq._core.QFT, conds: collections.abc.Sequence[tuple[typing.SupportsInt | typing.SupportsIndex, typing.SupportsInt | typing.SupportsIndex]]) -> pysparq._core.QFT

Condition this operation on a specific bit position.

Calling this method replaces prior bit conditions. Pass the list-of-pairs overload to require several bits simultaneously; conditions of different kinds are combined with logical AND.

Parameters:
  • cond – Register name (str) or ID (int).

  • pos – Bit position to check (0-indexed).

  • conds – List of (register, position) pairs.

Returns:

Self, for method chaining.

conditioned_by_nonzeros(*args, **kwargs)

Overloaded function.

  1. conditioned_by_nonzeros(self: pysparq._core.QFT, cond: str) -> pysparq._core.QFT

Condition this operation on registers with nonzero values.

Calling this method replaces prior nonzero-register conditions. Pass the list overload to require several registers simultaneously; conditions of different kinds are combined with logical AND.

Parameters:
  • cond – Register name (str) or ID (int) to condition on.

  • conds – List of register names or IDs for multi-condition.

Returns:

Self, for method chaining.

Example

op.conditioned_by_nonzeros(‘control_reg’)(state)

  1. conditioned_by_nonzeros(self: pysparq._core.QFT, conds: collections.abc.Sequence[str]) -> pysparq._core.QFT

Condition this operation on registers with nonzero values.

Calling this method replaces prior nonzero-register conditions. Pass the list overload to require several registers simultaneously; conditions of different kinds are combined with logical AND.

Parameters:
  • cond – Register name (str) or ID (int) to condition on.

  • conds – List of register names or IDs for multi-condition.

Returns:

Self, for method chaining.

Example

op.conditioned_by_nonzeros(‘control_reg’)(state)

  1. conditioned_by_nonzeros(self: pysparq._core.QFT, cond: typing.SupportsInt | typing.SupportsIndex) -> pysparq._core.QFT

Condition this operation on registers with nonzero values.

Calling this method replaces prior nonzero-register conditions. Pass the list overload to require several registers simultaneously; conditions of different kinds are combined with logical AND.

Parameters:
  • cond – Register name (str) or ID (int) to condition on.

  • conds – List of register names or IDs for multi-condition.

Returns:

Self, for method chaining.

Example

op.conditioned_by_nonzeros(‘control_reg’)(state)

  1. conditioned_by_nonzeros(self: pysparq._core.QFT, conds: collections.abc.Sequence[typing.SupportsInt | typing.SupportsIndex]) -> pysparq._core.QFT

Condition this operation on registers with nonzero values.

Calling this method replaces prior nonzero-register conditions. Pass the list overload to require several registers simultaneously; conditions of different kinds are combined with logical AND.

Parameters:
  • cond – Register name (str) or ID (int) to condition on.

  • conds – List of register names or IDs for multi-condition.

Returns:

Self, for method chaining.

Example

op.conditioned_by_nonzeros(‘control_reg’)(state)

conditioned_by_value(*args, **kwargs)

Overloaded function.

  1. conditioned_by_value(self: pysparq._core.QFT, cond: str, pos: typing.SupportsInt | typing.SupportsIndex) -> pysparq._core.QFT

Condition this operation on registers holding a specific value.

Calling this method replaces prior value conditions. Pass the list-of-pairs overload to require several values simultaneously; conditions of different kinds are combined with logical AND.

Parameters:
  • cond – Register name (str) or ID (int).

  • pos – Value to match.

  • conds – List of (register, value) pairs.

Returns:

Self, for method chaining.

  1. conditioned_by_value(self: pysparq._core.QFT, conds: collections.abc.Sequence[tuple[str, typing.SupportsInt | typing.SupportsIndex]]) -> pysparq._core.QFT

Condition this operation on registers holding a specific value.

Calling this method replaces prior value conditions. Pass the list-of-pairs overload to require several values simultaneously; conditions of different kinds are combined with logical AND.

Parameters:
  • cond – Register name (str) or ID (int).

  • pos – Value to match.

  • conds – List of (register, value) pairs.

Returns:

Self, for method chaining.

  1. conditioned_by_value(self: pysparq._core.QFT, cond: typing.SupportsInt | typing.SupportsIndex, pos: typing.SupportsInt | typing.SupportsIndex) -> pysparq._core.QFT

Condition this operation on registers holding a specific value.

Calling this method replaces prior value conditions. Pass the list-of-pairs overload to require several values simultaneously; conditions of different kinds are combined with logical AND.

Parameters:
  • cond – Register name (str) or ID (int).

  • pos – Value to match.

  • conds – List of (register, value) pairs.

Returns:

Self, for method chaining.

  1. conditioned_by_value(self: pysparq._core.QFT, conds: collections.abc.Sequence[tuple[typing.SupportsInt | typing.SupportsIndex, typing.SupportsInt | typing.SupportsIndex]]) -> pysparq._core.QFT

Condition this operation on registers holding a specific value.

Calling this method replaces prior value conditions. Pass the list-of-pairs overload to require several values simultaneously; conditions of different kinds are combined with logical AND.

Parameters:
  • cond – Register name (str) or ID (int).

  • pos – Value to match.

  • conds – List of (register, value) pairs.

Returns:

Self, for method chaining.

dag(self: pysparq._core.QFT, state: pysparq._core.SparseState) → None

Apply the adjoint (inverse) of this operation.

Parameters:

state – The quantum state to operate on.

Note: Only available for self-adjoint operators.

Operation: Performs the quantum Fourier transform on an integer register

Type constraints: UnsignedInteger or SignedInteger

Bit constraints: The register size determines the transform dimension

import pysparq as ps

ps.System.clear()

# 4-bit register
ps.System.add_register("q", ps.UnsignedInteger, 4)

state = ps.SparseState()

# Initialize to the basis state |1⟩
ps.Init_Unsafe("q", 1)(state)

print("Initial state:")
ps.pprint(state)
# |q=1⟩ : (1+0j)

# Apply the QFT
op = ps.QFT("q")
op(state)

print("\nAfter QFT:")
ps.pprint(state)
# Produces a superposition with uniformly distributed phases

# Undo (inverse transform)
op.dag(state)

InverseQFT

class pysparq.InverseQFT

Inverse Quantum Fourier Transform on a register.

Applies the inverse QFT to transform from Fourier basis back to computational basis.

Parameters:

reg_name – Name of the register (str) or register ID (int).

clear_control_all_ones(self: pysparq._core.InverseQFT) → None

Clear all control conditions of the specified type.

clear_control_by_bit(self: pysparq._core.InverseQFT) → None

Clear all control conditions of the specified type.

clear_control_by_value(self: pysparq._core.InverseQFT) → None

Clear all control conditions of the specified type.

clear_control_nonzeros(self: pysparq._core.InverseQFT) → None

Clear all control conditions of the specified type.

property condition_variable_all_ones
property condition_variable_by_bit
property condition_variable_by_value
property condition_variable_nonzeros
conditioned_by_all_ones(*args, **kwargs)

Overloaded function.

  1. conditioned_by_all_ones(self: pysparq._core.InverseQFT, cond: str) -> pysparq._core.InverseQFT

Condition this operation on registers where all bits are 1.

Calling this method replaces prior all-ones conditions. Pass the list overload to require several registers simultaneously; conditions of different kinds are combined with logical AND.

Parameters:
  • cond – Register name (str) or ID (int) to condition on.

  • conds – List of register names or IDs for multi-condition.

Returns:

Self, for method chaining.

  1. conditioned_by_all_ones(self: pysparq._core.InverseQFT, conds: collections.abc.Sequence[str]) -> pysparq._core.InverseQFT

Condition this operation on registers where all bits are 1.

Calling this method replaces prior all-ones conditions. Pass the list overload to require several registers simultaneously; conditions of different kinds are combined with logical AND.

Parameters:
  • cond – Register name (str) or ID (int) to condition on.

  • conds – List of register names or IDs for multi-condition.

Returns:

Self, for method chaining.

  1. conditioned_by_all_ones(self: pysparq._core.InverseQFT, cond: typing.SupportsInt | typing.SupportsIndex) -> pysparq._core.InverseQFT

Condition this operation on registers where all bits are 1.

Calling this method replaces prior all-ones conditions. Pass the list overload to require several registers simultaneously; conditions of different kinds are combined with logical AND.

Parameters:
  • cond – Register name (str) or ID (int) to condition on.

  • conds – List of register names or IDs for multi-condition.

Returns:

Self, for method chaining.

  1. conditioned_by_all_ones(self: pysparq._core.InverseQFT, conds: collections.abc.Sequence[typing.SupportsInt | typing.SupportsIndex]) -> pysparq._core.InverseQFT

Condition this operation on registers where all bits are 1.

Calling this method replaces prior all-ones conditions. Pass the list overload to require several registers simultaneously; conditions of different kinds are combined with logical AND.

Parameters:
  • cond – Register name (str) or ID (int) to condition on.

  • conds – List of register names or IDs for multi-condition.

Returns:

Self, for method chaining.

conditioned_by_bit(*args, **kwargs)

Overloaded function.

  1. conditioned_by_bit(self: pysparq._core.InverseQFT, cond: str, pos: typing.SupportsInt | typing.SupportsIndex) -> pysparq._core.InverseQFT

Condition this operation on a specific bit position.

Calling this method replaces prior bit conditions. Pass the list-of-pairs overload to require several bits simultaneously; conditions of different kinds are combined with logical AND.

Parameters:
  • cond – Register name (str) or ID (int).

  • pos – Bit position to check (0-indexed).

  • conds – List of (register, position) pairs.

Returns:

Self, for method chaining.

  1. conditioned_by_bit(self: pysparq._core.InverseQFT, conds: collections.abc.Sequence[tuple[str, typing.SupportsInt | typing.SupportsIndex]]) -> pysparq._core.InverseQFT

Condition this operation on a specific bit position.

Calling this method replaces prior bit conditions. Pass the list-of-pairs overload to require several bits simultaneously; conditions of different kinds are combined with logical AND.

Parameters:
  • cond – Register name (str) or ID (int).

  • pos – Bit position to check (0-indexed).

  • conds – List of (register, position) pairs.

Returns:

Self, for method chaining.

  1. conditioned_by_bit(self: pysparq._core.InverseQFT, cond: typing.SupportsInt | typing.SupportsIndex, pos: typing.SupportsInt | typing.SupportsIndex) -> pysparq._core.InverseQFT

Condition this operation on a specific bit position.

Calling this method replaces prior bit conditions. Pass the list-of-pairs overload to require several bits simultaneously; conditions of different kinds are combined with logical AND.

Parameters:
  • cond – Register name (str) or ID (int).

  • pos – Bit position to check (0-indexed).

  • conds – List of (register, position) pairs.

Returns:

Self, for method chaining.

  1. conditioned_by_bit(self: pysparq._core.InverseQFT, conds: collections.abc.Sequence[tuple[typing.SupportsInt | typing.SupportsIndex, typing.SupportsInt | typing.SupportsIndex]]) -> pysparq._core.InverseQFT

Condition this operation on a specific bit position.

Calling this method replaces prior bit conditions. Pass the list-of-pairs overload to require several bits simultaneously; conditions of different kinds are combined with logical AND.

Parameters:
  • cond – Register name (str) or ID (int).

  • pos – Bit position to check (0-indexed).

  • conds – List of (register, position) pairs.

Returns:

Self, for method chaining.

conditioned_by_nonzeros(*args, **kwargs)

Overloaded function.

  1. conditioned_by_nonzeros(self: pysparq._core.InverseQFT, cond: str) -> pysparq._core.InverseQFT

Condition this operation on registers with nonzero values.

Calling this method replaces prior nonzero-register conditions. Pass the list overload to require several registers simultaneously; conditions of different kinds are combined with logical AND.

Parameters:
  • cond – Register name (str) or ID (int) to condition on.

  • conds – List of register names or IDs for multi-condition.

Returns:

Self, for method chaining.

Example

op.conditioned_by_nonzeros(‘control_reg’)(state)

  1. conditioned_by_nonzeros(self: pysparq._core.InverseQFT, conds: collections.abc.Sequence[str]) -> pysparq._core.InverseQFT

Condition this operation on registers with nonzero values.

Calling this method replaces prior nonzero-register conditions. Pass the list overload to require several registers simultaneously; conditions of different kinds are combined with logical AND.

Parameters:
  • cond – Register name (str) or ID (int) to condition on.

  • conds – List of register names or IDs for multi-condition.

Returns:

Self, for method chaining.

Example

op.conditioned_by_nonzeros(‘control_reg’)(state)

  1. conditioned_by_nonzeros(self: pysparq._core.InverseQFT, cond: typing.SupportsInt | typing.SupportsIndex) -> pysparq._core.InverseQFT

Condition this operation on registers with nonzero values.

Calling this method replaces prior nonzero-register conditions. Pass the list overload to require several registers simultaneously; conditions of different kinds are combined with logical AND.

Parameters:
  • cond – Register name (str) or ID (int) to condition on.

  • conds – List of register names or IDs for multi-condition.

Returns:

Self, for method chaining.

Example

op.conditioned_by_nonzeros(‘control_reg’)(state)

  1. conditioned_by_nonzeros(self: pysparq._core.InverseQFT, conds: collections.abc.Sequence[typing.SupportsInt | typing.SupportsIndex]) -> pysparq._core.InverseQFT

Condition this operation on registers with nonzero values.

Calling this method replaces prior nonzero-register conditions. Pass the list overload to require several registers simultaneously; conditions of different kinds are combined with logical AND.

Parameters:
  • cond – Register name (str) or ID (int) to condition on.

  • conds – List of register names or IDs for multi-condition.

Returns:

Self, for method chaining.

Example

op.conditioned_by_nonzeros(‘control_reg’)(state)

conditioned_by_value(*args, **kwargs)

Overloaded function.

  1. conditioned_by_value(self: pysparq._core.InverseQFT, cond: str, pos: typing.SupportsInt | typing.SupportsIndex) -> pysparq._core.InverseQFT

Condition this operation on registers holding a specific value.

Calling this method replaces prior value conditions. Pass the list-of-pairs overload to require several values simultaneously; conditions of different kinds are combined with logical AND.

Parameters:
  • cond – Register name (str) or ID (int).

  • pos – Value to match.

  • conds – List of (register, value) pairs.

Returns:

Self, for method chaining.

  1. conditioned_by_value(self: pysparq._core.InverseQFT, conds: collections.abc.Sequence[tuple[str, typing.SupportsInt | typing.SupportsIndex]]) -> pysparq._core.InverseQFT

Condition this operation on registers holding a specific value.

Calling this method replaces prior value conditions. Pass the list-of-pairs overload to require several values simultaneously; conditions of different kinds are combined with logical AND.

Parameters:
  • cond – Register name (str) or ID (int).

  • pos – Value to match.

  • conds – List of (register, value) pairs.

Returns:

Self, for method chaining.

  1. conditioned_by_value(self: pysparq._core.InverseQFT, cond: typing.SupportsInt | typing.SupportsIndex, pos: typing.SupportsInt | typing.SupportsIndex) -> pysparq._core.InverseQFT

Condition this operation on registers holding a specific value.

Calling this method replaces prior value conditions. Pass the list-of-pairs overload to require several values simultaneously; conditions of different kinds are combined with logical AND.

Parameters:
  • cond – Register name (str) or ID (int).

  • pos – Value to match.

  • conds – List of (register, value) pairs.

Returns:

Self, for method chaining.

  1. conditioned_by_value(self: pysparq._core.InverseQFT, conds: collections.abc.Sequence[tuple[typing.SupportsInt | typing.SupportsIndex, typing.SupportsInt | typing.SupportsIndex]]) -> pysparq._core.InverseQFT

Condition this operation on registers holding a specific value.

Calling this method replaces prior value conditions. Pass the list-of-pairs overload to require several values simultaneously; conditions of different kinds are combined with logical AND.

Parameters:
  • cond – Register name (str) or ID (int).

  • pos – Value to match.

  • conds – List of (register, value) pairs.

Returns:

Self, for method chaining.

Operation: Performs the inverse quantum Fourier transform on an integer register

Purpose: Undo a QFT or extract phase information

ps.System.clear()

ps.System.add_register("q", ps.UnsignedInteger, 4)

state = ps.SparseState()

# Create a superposition
ps.Hadamard_Int_Full("q")(state)

# Apply the inverse QFT
ps.InverseQFT("q")(state)

# The inverse QFT and the QFT are inverses of each other
ps.QFT("q")(state)

—

Use Cases

Period Finding (Shor’s Algorithm)

import pysparq as ps
import numpy as np

ps.System.clear()

# Phase register
n_bits = 8
ps.System.add_register("phase", ps.UnsignedInteger, n_bits)

state = ps.SparseState()

# Simulate a phase estimation result (assume period r = 5)
# |phase⟩ contains periodic phase information
ps.Hadamard_Int_Full("phase")(state)

# Apply the inverse QFT to extract the period
ps.InverseQFT("phase")(state)

# Measure the phase register to determine the period
# ...

Phase Estimation

# Assume a phase estimation circuit already exists
# The inverse QFT converts phase information into the computational basis

# 1. Prepare the initial state
ps.Hadamard_Int_Full("estimate")(state)

# 2. Apply the controlled-unitary operation
# controlled_U(...)(state)

# 3. Extract the phase with the inverse QFT
ps.InverseQFT("estimate")(state)

# 4. Measure to obtain the phase estimate
# ...

Quantum Signal Processing

# The QFT is used for frequency-domain analysis
ps.System.add_register("signal", ps.UnsignedInteger, 16)

state = ps.SparseState()
# Load the signal data...

# Transform to the frequency domain
ps.QFT("signal")(state)

# Frequency-domain processing...

# Transform back to the time domain
ps.InverseQFT("signal")(state)

Performance Considerations

  • Number of basis states: When the QFT acts on a superposition, it produces \(2^n\) basis states

  • Memory limits: For large n (e.g. n > 20), memory problems may arise

  • Sparse optimization: PySparQ stores only the basis states with nonzero amplitudes, which can be efficient for certain inputs

Note

For superpositions that are not complete, the sparse implementation of the QFT can be more efficient than full-state simulation.

Practical Examples

Full QFT + InverseQFT Cycle

import pysparq as ps
import numpy as np

ps.System.clear()

n = 3
ps.System.add_register("q", ps.UnsignedInteger, n)

state = ps.SparseState()

# Initial value
initial_value = 5
ps.Init_Unsafe("q", initial_value)(state)

print("Initial state:")
ps.pprint(state)

# QFT
ps.QFT("q")(state)
print("\nAfter QFT:")
ps.pprint(state)

# InverseQFT
ps.InverseQFT("q")(state)
print("\nAfter InverseQFT:")
ps.pprint(state)
# Restored to |q=5⟩