Fix issues raised by pylint (#205)

* all pylint issues in tests/

* state_props:eof

* python install

* pylint userwarning

* remove exception and from future annotations

* import lines

* matrix_props/is_positive_def

* fixes Make parameter typing consistent in doc examples #209

* state_metrics/fid_of_sep

* channel_props/is_trace_preserving

* nonlocalgame

* useless suppression

* quick fixes

* all redefinition disabled

* all unsubscriptable object

* possibly unbalanced tuples

* unnecessary-list-index-lookup

* black reformatter no pylint issues

* black reformatter pylint issues

* test useless supression

* Fixes #205 (comment)

* Fixes #205 (comment)

* Fixes #205 (comment)

* fixes #205 (comment) + black reformatter

.pylintrc

@@ -54,7 +54,7 @@ unsafe-load-any-extension=no
 extension-pkg-allow-list=
 
 # Minimum supported python version
-py-version = 3.7.2
+py-version = 3.10
 
 # Control the amount of potential inferred values when inferring a single
 # object. This can help the performance when dealing with large functions or
@@ -204,7 +204,7 @@ redefining-builtins-modules=six.moves,past.builtins,future.builtins,builtins,io
 [FORMAT]
 
 # Maximum number of characters on a single line.
-max-line-length=100
+max-line-length=120
 
 # Regexp for a line that is allowed to be longer than the limit.
 ignore-long-lines=^\s*(# )?<?https?://\S+>?$
@@ -513,7 +513,7 @@ preferred-modules=
 
 # Exceptions that will emit a warning when being caught. Defaults to
 # "Exception"
-overgeneral-exceptions=Exception
+overgeneral-exceptions=builtins.Exception
 
 
 [TYPING]

pyproject.toml

@@ -36,7 +36,7 @@ keywords = ["quantum information", "quantum computing", "nonlocal games"]
 
 
 [tool.poetry.dependencies]
-python = "^3.9"
+python = "^3.10"
 cvxpy = "^1.2.1"
 cvxopt = "^1.2.5"
 numpy = "^1.19.4"

setup.py

@@ -3,7 +3,7 @@
 with open("README.md", "r") as fh:
     long_description = fh.read()
 
-requirements = ["cvx", "cvxpy", "numpy", "picos", "scipy", "scikit-image"]
+requirements = ["cvx", "cvxpy", "cvxopt", "numpy", "picos", "scipy", "scikit-image", "pytest", "pytest-cov"]
 
 setuptools.setup(
     name="toqito",
@@ -23,7 +23,7 @@
         "License :: OSI Approved :: BSD License",
         "Operating System :: OS Independent",
     ],
-    python_requires=">=3.9",
+    python_requires=">=3.10",
     install_requires=requirements,
     test_suite="tests",
 )

tests/test_channel_metrics/test_completely_bounded_trace_norm.py

@@ -16,7 +16,7 @@ def test_cb_trace_norm_unitaries_channel():
     """The diamond norm of phi = id- U id U* is the diameter of the smallest circle that contains the eigenvalues of U."""
     U = 1 / np.sqrt(2) * np.array([[1, 1], [-1, 1]])  # Hadamard gate
     phi = kraus_to_choi([[np.eye(2), np.eye(2)], [U, -U]])
-    lam, eigv = np.linalg.eig(U)
+    lam, _ = np.linalg.eig(U)
     dist = np.abs(lam[:, None] - lam[None, :])  # all to all distance
     diameter = np.max(dist)
     np.testing.assert_equal(

tests/test_channel_ops/test_dual_channel.py

@@ -75,8 +75,8 @@ def test_dual_channel_choi_dims():
 
 def test_dual_channel_nonsquare_matrix():
     """Dual of a channel that transposes 3x2 matrices."""
-    choi = swap_operator([2, 3])
-    choi_dual = dual_channel(choi, dims=[[3, 2], [2, 3]])
+    choi1 = swap_operator([2, 3])
+    choi_dual = dual_channel(choi1, dims=[[3, 2], [2, 3]])
     expected_choi_dual = np.array(
         [
             [1, 0, 0, 0, 0, 0],

tests/test_nonlocal_games/test_xor_game.py

@@ -3,7 +3,6 @@
 
 import numpy as np
 
-from toqito.nonlocal_games.nonlocal_game import NonlocalGame
 from toqito.nonlocal_games.xor_game import XORGame
 
 

toqito/channel_metrics/completely_bounded_trace_norm.py

@@ -44,7 +44,7 @@ def completely_bounded_trace_norm(phi: np.ndarray) -> float:
     if is_quantum_channel(phi):
         return 1
 
-    elif is_completely_positive(phi):
+    if is_completely_positive(phi):
         v = apply_channel(np.eye(dim_ly), dual_channel(phi))
         return trace_norm(v)
 

toqito/channel_metrics/fidelity_of_separability.py

@@ -14,7 +14,11 @@
 
 
 def fidelity_of_separability(
-    psi: np.ndarray, psi_dims: list[int], k: int = 1, verbosity_option=2, solver_option="cvxopt"
+    psi: np.ndarray,
+    psi_dims: list[int],
+    k: int = 1,
+    verbosity_option: int = 2,
+    solver_option: str = "cvxopt",
 ) -> float:
     r"""
     Define the first benchmark introduced in Appendix I of [Phil23.1]_ .
@@ -113,28 +117,25 @@ def fidelity_of_separability(
         “The Theory of Quantum Information”
         Cambridge University Press, 2018
 
-    Args:
-        psi: the density matrix for the tripartite state of interest psi_{BAR}
-        psi_dims: the dimensions of System A, B, & R in
+    
+    :param psi: the density matrix for the tripartite state of interest psi_{BAR}
+    :param psi_dims: the dimensions of System A, B, & R in
             the input state density matrix. It is assumed that the first
             quantity in this list is the dimension of System B.
-        k: value for k-extendibility.
-        verbosity_option: Parameter option for `picos`. Default value is `verbosity = 2`.
-            For more info, visit https://picos-api.gitlab.io/picos/api/picos.modeling.options.html#option-verbosity
-        solver_option: Optimization option for `picos` solver. Default option is `solver_option="cvxopt"`
-            For more info, visit https://picos-api.gitlab.io/picos/api/picos.modeling.options.html#option-solver
-    Raises:
-        AssertionError:
-            * If the provided dimensions are not for a tripartite density
-            matrix.
-        TypeError:
-            * If the matrix is not a density matrix (square matrix that is \n
-            PSD with trace 1).
-        TypeError:
-            * If the input state is a mixed state.
-    Returns:
-        Optimized value of the SDP when maximized over a set of linear
-        operators subject to some constraints.
+    :param k: value for k-extendibility.
+    :param verbosity_option: Parameter option for `picos`. Default value is 
+        `verbosity = 2`. For more info, visit
+        https://picos-api.gitlab.io/picos/api/picos.modeling.options.html#option-verbosity.
+    :param solver_option: Optimization option for `picos` solver. Default option is 
+        `solver_option="cvxopt"`. For more info, visit 
+        https://picos-api.gitlab.io/picos/api/picos.modeling.options.html#option-solver.
+    :raises AssertionError: If the provided dimensions are not for a tripartite density matrix.
+    :raises ValueError: If the matrix is not a density matrix (square matrix that
+        is PSD with trace 1).
+    :raises ValueError: the input state is entangled.
+    :raises ValueError: the input state is a mixed state.
+    :return: Optimized value of the SDP when maximized over a set of linear
+        operators subject to some constraints.  
     """
     if not is_density(psi):
         raise ValueError("Provided input state is not a density matrix.")

toqito/channel_ops/dual_channel.py

@@ -4,7 +4,6 @@
 import numpy as np
 
 from toqito.helper import channel_dim
-from toqito.matrix_props import is_square
 from toqito.perms import swap
 
 

toqito/channel_props/is_trace_preserving.py

@@ -100,9 +100,8 @@ def is_trace_preserving(
         k_r = np.concatenate(phi_r, axis=0)
 
         mat = k_l.conj().T @ k_r
+    elif dim is None:
+        mat = partial_trace(phi, [sys - 1])
     else:
-        if dim == None:
-            mat = partial_trace(phi, [sys - 1])
-        else:
-            mat = partial_trace(phi, [sys - 1], dim)
+        mat = partial_trace(phi, [sys - 1], dim)
     return is_identity(np.array(mat), rtol=rtol, atol=atol)

toqito/channels/partial_trace.py

@@ -12,7 +12,7 @@
 
 def partial_trace(
     input_mat: np.ndarray | Variable,
-    sys: int | list[int] = [1],
+    sys: int | list[int] = None,
     dim: int | list[int] = None,
 ) -> np.ndarray | Expression:
     r"""
@@ -130,6 +130,9 @@ def partial_trace(
                 equal.
     :return: The partial trace of matrix :code:`input_mat`.
     """
+    if not isinstance(sys, int):
+        if sys is None:
+            sys = [1]
     # If the input matrix is a CVX variable for an SDP, we convert it to a numpy array,
     # perform the partial trace, and convert it back to a CVX variable.
     if isinstance(input_mat, Variable):
@@ -145,10 +148,8 @@ def partial_trace(
     if isinstance(dim, list):
         dim = np.array(dim)
 
-    num_sys = len(dim)
-
     # Allow the user to enter a single number for dim.
-    if num_sys == 1:
+    if (num_sys := len(dim)) == 1:
         dim = np.array([dim[0], len(input_mat) / dim[0]])
         if np.abs(dim[1] - np.round(dim[1])) >= 2 * len(input_mat) * np.finfo(float).eps:
             raise ValueError(
@@ -166,7 +167,7 @@ def partial_trace(
             for idx in sys:
                 prod_dim_sys *= dim[idx]
     elif isinstance(sys, int):
-        prod_dim_sys = np.prod(dim[sys])
+        prod_dim_sys = np.prod(dim[sys])  # pylint: disable=redefined-variable-type
     else:
         raise ValueError(
             "Invalid: The variable `sys` must either be of type int or of a list of ints."
@@ -176,7 +177,7 @@ def partial_trace(
     sub_sys_vec = prod_dim * np.ones(int(sub_prod)) / sub_prod
 
     if isinstance(sys, list):
-        sys = np.array(sys)
+        sys = np.array(sys)  # pylint: disable=redefined-variable-type
     if isinstance(sys, int):
         sys = np.array([sys])
     set_diff = list(set(list(range(1, num_sys + 1))) - set(sys + 1))

toqito/channels/partial_transpose.py

@@ -12,7 +12,7 @@
 
 def partial_transpose(
     rho: np.ndarray | Variable,
-    sys: list[int] | np.ndarray | int = [1],
+    sys: list[int] | np.ndarray | int = None,
     dim: list[int] | np.ndarray = None,
 ) -> np.ndarray | Expression:
     r"""Compute the partial transpose of a matrix [WikPtrans]_.
@@ -113,6 +113,9 @@ def partial_transpose(
     :raises ValueError: If matrix dimensions are not square.
     :returns: The partial transpose of matrix :code:`rho`.
     """
+    if not isinstance(sys, int):
+        if sys is None:
+            sys = [1]
     # If the input matrix is a CVX variable for an SDP, we convert it to a
     # numpy array, perform the partial transpose, and convert it back to a CVX
     # variable.
@@ -131,13 +134,12 @@ def partial_transpose(
     if isinstance(dim, list):
         dim = np.array(dim)
     if isinstance(sys, list):
-        sys = np.array(sys)
+        sys = np.array(sys)  # pylint: disable=redefined-variable-type
     if isinstance(sys, int):
         sys = np.array([sys])
 
-    num_sys = max(dim.shape)
     # Allow the user to enter a single number for dim.
-    if num_sys == 1:
+    if (num_sys := max(dim.shape)) == 1:
         dim = np.array([dim, list(rho.shape)[0] / dim])
         if np.abs(dim[1] - np.round(dim[1]))[0] >= 2 * list(rho.shape)[0] * np.finfo(float).eps:
             raise ValueError(

toqito/channels/realignment.py

@@ -79,7 +79,7 @@ def realignment(input_mat: np.ndarray, dim: int | list[int] = None) -> np.ndarra
         dim = np.array([dim, dim])
 
     dim_x = np.array([[dim[0][1], dim[0][0]], [dim[1][0], dim[1][1]]])
-    dim_x = np.int_(dim_x)
+    dim_x = np.int_(dim_x)  # pylint: disable=redefined-variable-type
     dim_y = np.array([[dim[1][0], dim[0][0]], [dim[0][1], dim[1][1]]])
 
     x_tmp = swap(input_mat, [1, 2], dim, True)

toqito/helper/npa_hierarchy.py

@@ -124,7 +124,7 @@ def _get_nonlocal_game_params(
     return a_out, a_in, b_out, b_in
 
 
-def npa_constraints(
+def npa_constraints(  # pylint: disable=too-many-locals
     assemblage: dict[tuple[int, int], cvxpy.Variable], k: int | str = 1, referee_dim: int = 1
 ) -> list[cvxpy.constraints.constraint.Constraint]:
     r"""

toqito/matrices/pauli.py

@@ -97,7 +97,7 @@ def pauli(
             pauli_mat = np.identity(2)
 
         if is_sparse:
-            pauli_mat = sparse.csr_matrix(pauli_mat)
+            pauli_mat = sparse.csr_matrix(pauli_mat)  # pylint: disable=redefined-variable-type
 
         return pauli_mat
 

toqito/matrix_ops/inner_product.py

@@ -33,7 +33,8 @@ def inner_product(v1: np.ndarray, v2: np.ndarray) -> float:
         https://en.wikipedia.org/wiki/Inner_product_space
 
     :raises ValueError: Vector dimensions are mismatched.
-    :param args: v1 and v2, both vectors of dimenstions :math:`(n,1)` where :math:`n>1`.
+    :param v1: v1 and v2, both vectors of dimenstions :math:`(n,1)` where :math:`n>1`.
+    :param v2: v1 and v2, both vectors of dimenstions :math:`(n,1)` where :math:`n>1`.
     :return: The computed inner product.
     """
     # Check for dimensional validity

toqito/matrix_ops/outer_product.py

@@ -35,7 +35,8 @@ def outer_product(v1: np.ndarray, v2: np.ndarray) -> np.ndarray:
         https://en.wikipedia.org/wiki/Outer_product
 
     :raises ValueError: Vector dimensions are mismatched.
-    :param args: v1 and v2, both vectors of dimensions :math:`(n,1)` where :math:`n>1`.
+    :param v1: v1 and v2, both vectors of dimenstions :math:`(n,1)` where :math:`n>1`.
+    :param v2: v1 and v2, both vectors of dimenstions :math:`(n,1)` where :math:`n>1`.
     :return: The computed outer product.
     """
     # Check for dimensional validity

toqito/matrix_ops/vectors_from_gram_matrix.py

@@ -3,14 +3,18 @@
 
 
 def vectors_from_gram_matrix(gram: np.ndarray) -> list[np.ndarray]:
-    """Obtain the corresponding ensemble of states from the Gram matrix."""
+    """Obtain the corresponding ensemble of states from the Gram matrix.
+
+    :param gram: Input Gram matrix.
+    :return: list of ensemble states
+    """
     dim = gram.shape[0]
     # If matrix is PD, can do Cholesky decomposition:
     try:
         decomp = np.linalg.cholesky(gram)
         return [decomp[i][:] for i in range(dim)]
     # Otherwise, need to do eigendecomposition:
-    except Exception:
+    except np.linalg.LinAlgError:
         print("Matrix is not positive semidefinite. Using eigendecomposition as alternative.")
         d, v = np.linalg.eig(gram)
         return [np.sqrt(np.diag(d)) @ v[i].conj().T for i in range(dim)]

toqito/matrix_props/__init__.py

@@ -18,6 +18,5 @@
 from toqito.matrix_props.majorizes import majorizes
 from toqito.matrix_props.sk_norm import sk_operator_norm
 from toqito.matrix_props.is_block_positive import is_block_positive
-from toqito.matrix_props.is_orthonormal import is_orthonormal
 from toqito.matrix_props.trace_norm import trace_norm
 from toqito.matrix_props.is_diagonally_dominant import is_diagonally_dominant

toqito/matrix_props/is_block_positive.py

@@ -68,7 +68,7 @@ def is_block_positive(
 
     # Allow the user to enter in a single integer for dimension.
     if isinstance(dim, int):
-        dim = np.array([dim, dim_xy / dim])
+        dim = np.array([dim, dim_xy / dim])  # pylint: disable=redefined-variable-type
         if np.abs(dim[1] - np.round(dim[1])) >= 2 * dim_xy * np.finfo(float).eps:
             raise ValueError(
                 "If `dim` is a scalar, it must evenly divide the length of the matrix."

toqito/matrix_props/is_positive_definite.py

@@ -1,10 +1,10 @@
 """Is matrix a positive definite matrix."""
 import numpy as np
 
-from toqito.matrix_props import is_hermitian
+from toqito.matrix_props import is_hermitian  # pylint: disable=unused-import
 
 
-def is_positive_definite(mat: np.ndarray, rtol: float = 1e-05, atol: float = 1e-08) -> bool:
+def is_positive_definite(mat: np.ndarray) -> bool:
     r"""
     Check if matrix is positive definite (PD) [WikPD]_.
 
@@ -54,8 +54,6 @@ def is_positive_definite(mat: np.ndarray, rtol: float = 1e-05, atol: float = 1e-
         https://en.wikipedia.org/wiki/Definiteness_of_a_matrix
 
     :param mat: Matrix to check.
-    :param rtol: The relative tolerance parameter (default 1e-05).
-    :param atol: The absolute tolerance parameter (default 1e-08).
     :return: Return :code:`True` if matrix is positive definite, and :code:`False` otherwise.
     """
     if np.array_equal(mat, mat.conj().T):