Pymbolic typing

.github/workflows/ci.yml

@@ -31,7 +31,7 @@ jobs:
         runs-on: ubuntu-latest
         strategy:
             matrix:
-                python-version: ['3.8', '3.x']
+                python-version: ['3.10', '3.x']
         steps:
         -   uses: actions/checkout@v4
         -

contrib/pov-nodes/nodal_points.py

@@ -47,7 +47,7 @@
         unit_to_barycentric(mp.warp_and_blend_nodes(3, n, node_tuples))
     ).T
 ]
-id_to_node = dict(list(zip(node_tuples, nodes)))
+id_to_node = dict(list(zip(node_tuples, nodes, strict=True)))
 
 
 def get_ball_radius(nid):

contrib/pov-nodes/pov.py

@@ -72,7 +72,7 @@ def __init__(self, name, args=(), opts=(), **kwargs):
 
         args = list(args)
         for i in range(len(args)):
-            if isinstance(args[i], (tuple, list, np.ndarray)):
+            if isinstance(args[i], tuple | list | np.ndarray):
                 args[i] = Vector(args[i])
 
         self.args = args
@@ -95,7 +95,7 @@ def write(self, file):
         for key, val in list(self.kwargs.items()):
             if val is None:
                 file.writeln(key)
-            elif isinstance(val, (tuple, list, np.ndarray)):
+            elif isinstance(val, tuple | list | np.ndarray):
                 val = Vector(*val)
                 file.writeln(f"{key} {val}")
             else:

examples/plot-basis.py

@@ -33,6 +33,7 @@
 for (i, j), basis_func in zip(
         gnitstam(p, dims),
         simplex_onb(dims, p),
+        strict=True,
         ):
 
     all_nodes.append([*plot_nodes, stretch_factor * i, stretch_factor * j])

modepy/matrices.py

@@ -24,7 +24,7 @@
 """
 
 
-from typing import Callable, Sequence
+from collections.abc import Callable, Sequence
 from warnings import warn
 
 import numpy as np

modepy/modal_decay.py

@@ -114,7 +114,7 @@ def create_decay_baseline(mode_number_vector: np.ndarray, n: int) -> np.ndarray:
     modal_coefficients[zeros] = 1  # irrelevant, just keeps log from NaNing
 
     # NOTE: mypy seems to be confused by the __itruediv__ argument types
-    modal_coefficients /= la.norm(modal_coefficients)   # type: ignore[misc]
+    modal_coefficients /= la.norm(modal_coefficients)   # type: ignore[arg-type,misc]
 
     return modal_coefficients
 

modepy/modes.py

@@ -26,14 +26,10 @@
 
 import math
 from abc import ABC, abstractmethod
+from collections.abc import Callable, Hashable, Iterable, Sequence
 from functools import partial, singledispatch
 from typing import (
     TYPE_CHECKING,
-    Callable,
-    Hashable,
-    Iterable,
-    Sequence,
-    Tuple,
     TypeVar,
 )
 
@@ -134,7 +130,9 @@ def _cse(expr, prefix):
 def _where(op_a, comp, op_b, then, else_):
     from pymbolic.primitives import Comparison, Expression, If
     if isinstance(op_a, Expression) or isinstance(op_b, Expression):
-        return If(Comparison(op_a, comp, op_b), then, else_)
+        return If(
+              Comparison(op_a, Comparison.name_to_operator[comp], op_b),
+              then, else_)
 
     import operator
     comp_op = getattr(operator, comp)
@@ -203,7 +201,7 @@ def jacobi(alpha: float, beta: float, n: int, x: RealValueT) -> RealValueT:
 
         bnew = -(alpha*alpha-beta*beta)/(h1*(h1+2.))
         pl.append(_cse(
-            (-aold*pl[i-1] + np.multiply(x-bnew, pl[i]))/anew,
+            (-aold*pl[i-1] + (x-bnew) * pl[i])/anew,
             prefix=f"jac_p{i+1}"))
         aold = anew
 
@@ -577,7 +575,7 @@ def __call__(self, x):
             # Likely we're evaluating symbolically.
             result = 1
 
-        for d, function in zip(self.dims_per_function, self.functions):
+        for d, function in zip(self.dims_per_function, self.functions, strict=True):
             result *= function(x[n:n + d])
             n += d
 
@@ -654,7 +652,7 @@ def __call__(self, x):
         n = 0
         for ider, derivative in enumerate(self.derivatives):
             f = 0
-            for iaxis, function in zip(self.dims_per_function, derivative):
+            for iaxis, function in zip(self.dims_per_function, derivative, strict=True):
                 components = function(x[f:f + iaxis])
 
                 if isinstance(components, tuple):
@@ -723,7 +721,7 @@ def symbolicize_function(
 # {{{ basis interface
 
 BasisFunctionType = Callable[[np.ndarray], np.ndarray]
-BasisGradientType = Callable[[np.ndarray], Tuple[np.ndarray, ...]]
+BasisGradientType = Callable[[np.ndarray], tuple[np.ndarray, ...]]
 
 
 class BasisNotOrthonormal(Exception):
@@ -973,7 +971,8 @@ def part_flat_tuple(iterable: Iterable[tuple[bool, Hashable]]
                 part_flat_tuple((flatten, umid[mid_index_i])
                       for flatten, umid, mid_index_i in zip(
                           is_all_singletons_with_int,
-                          underlying_mode_id_lists, mode_index_tuple))
+                          underlying_mode_id_lists, mode_index_tuple,
+                          strict=True))
                 for mode_index_tuple in self._mode_index_tuples)
 
     @property
@@ -998,7 +997,7 @@ def gradients(self) -> tuple[BasisGradientType, ...]:
                     tuple(
                         tuple(func[is_deriv][ibasis][mid_i]
                             for ibasis, (is_deriv, mid_i) in enumerate(
-                                zip(deriv_indicator_vec, mid)))
+                                zip(deriv_indicator_vec, mid, strict=True)))
                         for deriv_indicator_vec in wandering_element(self._nbases)),
                     dims_per_function=self._dims_per_basis)
                 for mid in self._mode_index_tuples)
@@ -1015,7 +1014,7 @@ def _orthonormal_basis_for_tp(
         raise ValueError("spatial dimensions of shape and space must match")
 
     bases = [orthonormal_basis_for_space(b, s)
-            for b, s in zip(space.bases, shape.bases)]
+            for b, s in zip(space.bases, shape.bases, strict=True)]
 
     return TensorProductBasis(
             bases,
@@ -1030,7 +1029,8 @@ def _basis_for_tp(space: TensorProductSpace, shape: TensorProductShape):
     if space.spatial_dim != shape.dim:
         raise ValueError("spatial dimensions of shape and space must match")
 
-    bases = [basis_for_space(b, s) for b, s in zip(space.bases, shape.bases)]
+    bases = [basis_for_space(b, s)
+        for b, s in zip(space.bases, shape.bases, strict=True)]
     return TensorProductBasis(
             bases,
             dims_per_basis=tuple(b.spatial_dim for b in space.bases))
@@ -1043,7 +1043,7 @@ def _monomial_basis_for_tp(space: TensorProductSpace, shape: TensorProductShape)
 
     bases = [
             monomial_basis_for_space(b, s)
-            for b, s in zip(space.bases, shape.bases)]
+            for b, s in zip(space.bases, shape.bases, strict=True)]
 
     return TensorProductBasis(
             bases,

modepy/nodes.py

@@ -62,8 +62,8 @@
 
 # }}}
 
+from collections.abc import Sequence
 from functools import partial, singledispatch
-from typing import Sequence
 
 import numpy as np
 import numpy.linalg as la
@@ -555,7 +555,7 @@ def _equispaced_nodes_for_tp(
 
     return tensor_product_nodes([
         equispaced_nodes_for_space(b, s)
-        for b, s in zip(space.bases, shape.bases)
+        for b, s in zip(space.bases, shape.bases, strict=True)
         ])
 
 
@@ -571,7 +571,7 @@ def _edge_clustered_nodes_for_tp(
 
     return tensor_product_nodes([
         edge_clustered_nodes_for_space(b, s)
-        for b, s in zip(space.bases, shape.bases)
+        for b, s in zip(space.bases, shape.bases, strict=True)
         ])
 
 

modepy/quadrature/__init__.py

@@ -50,9 +50,9 @@
 THE SOFTWARE.
 """
 
+from collections.abc import Callable, Iterable, Sequence
 from functools import singledispatch
 from numbers import Number
-from typing import Callable, Iterable, Sequence
 
 import numpy as np
 
@@ -79,7 +79,7 @@ def __gt__(self, other: object) -> bool:
         return bool(isinstance(other, Number))
 
     def __ge__(self, other: object) -> bool:
-        return bool(isinstance(other, (Number, _Inf)))
+        return bool(isinstance(other, Number | _Inf))
 
 
 inf = _Inf()
@@ -298,7 +298,7 @@ def _quadrature_for_tp(
     else:
         quad = TensorProductQuadrature([
             quadrature_for_space(sp, shp)
-            for sp, shp in zip(space.bases, shape.bases)
+            for sp, shp in zip(space.bases, shape.bases, strict=True)
             ])
 
     assert all(quad.exact_to >= getattr(s, "order", 0) for s in space.bases)

modepy/quadrature/grundmann_moeller.py

@@ -134,7 +134,8 @@ def __init__(self, order: int, dimension: int) -> None:
 
         dim_factor = 2**n
         for p, w in points_to_weights.items():
-            real_p = reduce(add, (a/b * v for (a, b), v in zip(p, vertices)))
+            real_p = reduce(add, (a/b * v
+                                for (a, b), v in zip(p, vertices, strict=True)))
             nodes.append(real_p)
             weights.append(dim_factor * w)
 

modepy/shapes.py

@@ -217,9 +217,10 @@
 
 import contextlib
 from abc import ABC, abstractmethod
+from collections.abc import Callable, Sequence
 from dataclasses import dataclass
 from functools import partial, singledispatch
-from typing import Any, Callable, Sequence
+from typing import Any
 
 import numpy as np
 
@@ -291,7 +292,7 @@ def face_normal(face: Face, normalize: bool = True) -> np.ndarray:
     from operator import xor as outerprod
 
     from pymbolic.geometric_algebra import MultiVector
-    surface_ps = reduce(outerprod, [
+    surface_ps: MultiVector = reduce(outerprod, [
         MultiVector(face_vertices[:, i+1] - face_vertices[:, 0])
         for i in range(face.dim)])
 

modepy/test/test_matrices.py

@@ -24,7 +24,7 @@
 """
 
 
-from typing import Tuple, cast
+from typing import cast
 
 import numpy as np
 import numpy.linalg as la
@@ -110,8 +110,8 @@ def test_tensor_product_diag_mass_matrix(shape: mp.Shape) -> None:
         # Note that gll_diag_mass_mat is not a good approximation of gll_ref_mass_mat
         # in the matrix norm sense!
 
-        for mid, func in zip(basis.mode_ids, basis.functions):
-            if max(cast(Tuple[int, ...], mid)) < order - 1:
+        for mid, func in zip(basis.mode_ids, basis.functions, strict=True):
+            if max(cast(tuple[int, ...], mid)) < order - 1:
                 err = np.abs(
                     gll_ref_mass_mat @ func(gll_quad.nodes)
                     - gll_diag_mass_mat @ func(gll_quad.nodes))

modepy/test/test_modes.py

@@ -178,7 +178,7 @@ def test_basis_grad(dim, shape_cls, order, basis_getter):
 
     from pytools import wandering_element
     from pytools.convergence import EOCRecorder
-    for bf, gradbf in zip(basis.functions, basis.gradients):
+    for bf, gradbf in zip(basis.functions, basis.gradients, strict=True):
         eoc_rec = EOCRecorder()
         for h in [1e-2, 1e-3]:
             r = mp.random_nodes_for_shape(shape, nnodes=1000, rng=rng)
@@ -246,7 +246,7 @@ def test_symbolic_basis(shape, order, basis_getter):
     rng = np.random.Generator(np.random.PCG64(17))
     r = mp.random_nodes_for_shape(shape, 10000, rng=rng)
 
-    for func, sym_func in zip(basis.functions, sym_basis):
+    for func, sym_func in zip(basis.functions, sym_basis, strict=True):
         strmap = MyStringifyMapper()
         s = strmap(sym_func)
         for name, val in strmap.cse_name_list:
@@ -276,7 +276,7 @@ def test_symbolic_basis(shape, order, basis_getter):
 
     sym_grad_basis = [mp.symbolicize_function(f, shape.dim) for f in basis.gradients]
 
-    for grad, sym_grad in zip(basis.gradients, sym_grad_basis):
+    for grad, sym_grad in zip(basis.gradients, sym_grad_basis, strict=True):
         strmap = MyStringifyMapper()
         s = strmap(sym_grad)
         for name, val in strmap.cse_name_list:
@@ -290,7 +290,7 @@ def test_symbolic_basis(shape, order, basis_getter):
             sym_val = (sym_val,)
             ref_val = (ref_val,)
 
-        for sv_i, rv_i in zip(sym_val, ref_val):
+        for sv_i, rv_i in zip(sym_val, ref_val, strict=True):
             ref_norm = la.norm(rv_i, np.inf)
             err = la.norm(sv_i-rv_i, np.inf)
             if ref_norm:

modepy/test/test_tools.py

@@ -580,7 +580,8 @@ def test_tensor_product_vdm_dim_by_dim(dim):
     x_r = reshape_array_for_tensor_product_space(space, x)
     vdm_dimbydim_x_r = x_r
 
-    for i, (subspace, subshape) in enumerate(zip(space.bases, shape.bases)):
+    for i, (subspace, subshape) in enumerate(
+                            zip(space.bases, shape.bases, strict=True)):
         subnodes = mp.edge_clustered_nodes_for_space(subspace, subshape)
         subbasis = mp.basis_for_space(subspace, subshape)
         subvdm = mp.vandermonde(subbasis.functions, subnodes)

pyproject.toml

@@ -1,7 +1,7 @@
 [build-system]
-build-backend = "setuptools.build_meta"
+build-backend = "hatchling.build"
 requires = [
-    "setuptools>=63",
+    "hatchling",
 ]
 
 [project]
@@ -13,7 +13,7 @@ license = { text = "MIT" }
 authors = [
     { name = "Andreas Kloeckner", email = "[email protected]" },
 ]
-requires-python = ">=3.8"
+requires-python = ">=3.10"
 classifiers = [
     "Development Status :: 3 - Alpha",
     "Intended Audience :: Developers",
@@ -32,7 +32,7 @@ classifiers = [
 ]
 dependencies = [
     "numpy",
-    "pymbolic",
+    "pymbolic>=2024.1",
     "pytools",
 ]
 
@@ -53,20 +53,6 @@ Documentation = "https://documen.tician.de/modepy"
 Homepage = "https://mathema.tician.de/software/modepy"
 Repository = "https://github.com/inducer/modepy"
 
-[tool.setuptools.packages.find]
-include = [
-    "modepy*",
-]
-
-[tool.setuptools.package-dir]
-# https://github.com/Infleqtion/client-superstaq/pull/715
-"" = "."
-
-[tool.setuptools.package-data]
-modepy = [
-    "py.typed",
-]
-
 [tool.ruff]
 preview = true