taulab 0.1.0

TAU physics-lab helpers for marimo notebooks

taulab

Physics-lab helpers for TAU coursework, built for marimo notebooks.

A rewrite of itayf-tau/taulab with uncertainty-propagating arithmetic, a stock instrument-uncertainty library (Keysight 34401A, DSO7012A, ruler/caliper), auto-seeded ODR, confidence-band plotting, and marimo-native display helpers.

Install

pip install taulab              # core
pip install taulab[marimo]      # + marimo display helpers

Quick example

import numpy as np
from taulab import fit_functions, odr_fit, Graph

x  = np.linspace(0, 1, 10)
y  = 2 * x + 1 + 0.02 * np.random.default_rng(0).standard_normal(10)
sx = np.full_like(x, 0.01)
sy = np.full_like(y, 0.02)

# Auto-seeded linear ODR — no init_values needed
res = odr_fit(fit_functions.linear, None, x, sx, y, sy,
              param_names=["intercept", "slope"])
print(res)                                              # metrics + params

fig, _ = Graph(res).plot_with_residuals(show_band=True)

In a marimo cell:

from taulab.display import fit_callout, params_table
mo.vstack([fit_callout(res), params_table(res, show_correlation=True)])

Highlights over upstream

• Uncertainty arithmetic on PhysicalSize via the uncertainties package.
• stats — resolution_sigma, sem, combine, weighted_mean, nsigma, drop_outliers.
• instruments — Keysight 34401A (resistance + DCV, auto-ranged), DSO7012A dual-cursor, ruler/caliper/micrometer, reading_lsd, column_lsd.
• ODR auto-seeded for linear / polynomial / constant fits.
• FitResult — cov, correlation_matrix, pulls, quality, extrapolate, params_dict.
• Graph.plot_with_residuals with a 1-σ confidence band from the parameter covariance.
• display — siunitx LaTeX formatters, fit_callout, params_table, results_table, uncertainty_budget, fit_metrics_accordion, chi_squared_badge, statistical_assessment_prose.

Full upstream import paths are preserved: from taulab.fit import odr_fit, fit_functions and from taulab.parse import parse_csv_folder both work unchanged.

License

MIT — see LICENSE.