SciNumTools3 for Python

SciNumTools3 provides unit-aware quantities and validated scientific configuration for Python. Use it to convert physical units, perform dimension-aware calculations, and load typed parameters whose units and constraints are checked at runtime.

The package includes:

• scinumtools3.puq for quantities, uncertainties, unit systems, and conversions
• scinumtools3.dip for typed, validated configuration files
• scinumtools3.api for formatted, command-style evaluation
• NumPy interoperability for array values

Installation

SciNumTools3 requires Python 3.9 or later and depends on NumPy.

pip install scinumtools3

Quick start

from scinumtools3.puq import Quantity

distance = Quantity(1.5, "km")
time = Quantity("2*min")

speed = (distance / time).convert("m*s-1")
print(speed)  # 12.5*m*s-1

Incompatible dimensions are rejected instead of being combined silently.

Quantities and units

Quantity accepts strings, Python numbers and sequences, and NumPy arrays. It supports unit-aware arithmetic, comparisons, uncertainties, and conversion between compatible units and unit systems.

PUEL (Physical Units Expression Language) expressions may contain scalar or array values, uncertainties, dimensions, constants, and units. For example:

from scinumtools3.puq import Quantity

lengths = Quantity("[2,3,4]*km")
measurement = Quantity("1.346591(30)e27*kg")

For conversions between unit systems, pass the target system and, when the dimensions are ambiguous between systems, the physical-quantity context:

from scinumtools3.puq import Quantity
from scinumtools3.puq.systems import ESU

current = Quantity("23*A").convert("statA", ESU, "I")

Input parameters

DIP parses DIPL (Dimensional Input Parameter Language) definitions from strings or files. DIPL provides typed parameters, units, constraints, dependencies, groups, lists, maps, and references.

Given parameters.dip:

simulation
  timestep float = 0.5 fs
    !condition ({.} > 0 fs)
  steps int = 1000000
    !condition ({.} > 1)
  duration float = ( {?simulation.timestep} * {?simulation.steps} ) fs

Parse it into an environment:

from pathlib import Path
from scinumtools3.dip import DIP

dip = DIP()
dip.add_file(Path("parameters.dip"))
env = dip.parse()

You can also add definitions directly:

dip = DIP()
dip.add_string("length float = 2.5 dm")
dip.add_string("count int = 42")
env = dip.parse()

Accessing nodes

There are two main ways to access nodes in an environment:

• Cursor: use env["path"] to inspect a known path or traverse groups and collections.
• Select: use env.select(...) to discover value nodes by path and tags and inspect snapshots of the results.

Cursor: inspect a known path

cursor = env["length"]
print(cursor.value, cursor.units, cursor.metadata.description)

A cursor accesses a path in the environment. Use cursor.elements() to traverse list elements and cursor.items() to traverse named items in a map. Cursor values are ordinary Python values: bool, int, float, str, or nested lists. Use cursor.to_numpy() to obtain a NumPy array. A cursor also exposes its units, shape, and child traversal methods. To convert a unit-bearing DIPL result, construct a quantity from its value and the string representation of its units:

from scinumtools3.puq import Quantity

length = env["length"]
length_cm = Quantity(length.value, length.units.to_string()).convert("cm")
print(length_cm)  # 25*cm

Select: discover and inspect nodes

Use select() to discover values and inspect their tags and metadata directly:

dip = DIP()
dip.add_string('''physics
  speed float = 2 m/s
    !tags ["export", "runtime", "hydro"]
    ?descr "Flow speed"
''')
env = dip.parse()
for node in env.select(
    "?physics.",
    tags_all=["export", "runtime"],
    tags_any=["hydro", "gravity"],
    tags_none=["internal", "deprecated"],
):
    print(node.name, node.value, node.tags, node.metadata.description)

The three filters combine with AND; omitted or empty filters impose no restriction. Only explicitly assigned tags match: children do not inherit their parent's tags. ? selects the whole environment, ?physics. selects a subtree including its value-bearing root and collection members, and ?physics.speed selects an exact node. Results retain fully qualified paths in environment order, with each node returned once. No matches returns an empty list.

Selected nodes are independent snapshots; selection does not modify the environment. Tags are returned as a separate Python list, and metadata remains valid while its Python wrapper is alive.

Additional request helpers remain available in the Python API reference.

Command-style API helpers

The scinumtools3.api.puq and scinumtools3.api.dip modules expose the same application-oriented operations as the snt command-line interface:

from scinumtools3.api.puq import PUQEval

command = PUQEval("1*cm + 2*dm")
print(command.execute())  # 21*cm

Use Quantity and DIP when you need typed Python objects; use the API helpers when you want formatted command-style output without starting a subprocess.

Documentation

• Python usage guide
• Python API reference
• PUEL specification
• DIPL specification
• GitHub repository
• Issue tracker

License

SciNumTools3 is distributed under the MIT License.