Examples#
The notebooks below walk through the core STAMP workflow, validate stereological
corrections, demonstrate the multi-state pipeline, and show MIPAR feature-measurement
analysis. Run them interactively with uv run jupyter lab.
Loading data#
stamp.io.load returns a single-column pd.DataFrame. The physical unit and
display label are stored in df.attrs["unit"] and df.attrs["label"].
All stamp.stats, stamp.stereo, and stamp.plot functions accept this
DataFrame directly — no manual unpacking needed:
from stamp.io import load
from stamp.stats import describe
from stamp.stereo import saltykov, two_step
from stamp.plot import distribution
ecds = load("grains.csv", column="ECD_um", unit="µm", label="Grain ECD")
stats = describe(ecds) # pd.DataFrame accepted directly
sal = saltykov(ecds, n_bins=12)
ts = two_step(ecds)
fig = distribution(ecds)
stamp.io.load_mipar_features also returns a pd.DataFrame (with all MIPAR
columns preserved). The stamp.pipeline functions (run, run_batch,
run_mipar) handle loading and type-conversion internally.
Notebooks#
01 — Quick Start#
01_quickstart.ipynb — end-to-end workflow for a single material state:
load measurements from a text file, compute descriptive statistics with
confidence intervals, fit a lognormal distribution, apply Saltykov / two-step
stereological correction, generate publication-ready figures, and export results
tables as CSV or LaTeX using stamp.export (§9 demonstrates default, Nature,
and custom journal styles).
02 — 2D → 3D Stereological Corrections#
02_simulation_validation.ipynb — Monte Carlo Wicksell validation: simulate
a synthetic lognormal 3-D grain population, generate 2-D cross-sections, apply
Saltykov and two-step corrections, and quantify recovery accuracy across a range
of sample sizes and bin counts.
03 — Multi-state Analysis#
03_multi_state_pipeline.ipynb — stamp.pipeline.run_batch applied to three
heat-treatment states stored as single batch CSV files. Demonstrates apparent (2-D)
vs stereologically corrected (3-D) geometric means side-by-side with ground-truth
reference lines.
04 — MIPAR Feature Measurements#
04_mipar_feature_analysis.ipynb — stamp.pipeline.run_mipar applied to
MIPAR feature-measurement CSVs containing multiple precipitate phases (M23C6,
MX ZPhase, Laves) across two material states (GOO220_52 vs GOO220_53). Shows
per-FOV ECD box plots for each phase rendered in Nature journal style (B&W,
hatch-differentiated boxes, 89 mm / 180 mm column widths) and exports the
summary statistics table as both CSV and a LaTeX booktabs table via
stamp.export.to_latex.
05 — MIPAR Image Measurements#
05_mipar_image_analysis.ipynb — stamp.io.load_mipar_image applied to
MIPAR batch image-measurement CSVs (one row per FOV, all phases as column
suffixes) across three material states (GOO220_51, GOO220_52, GOO220_53).
Demonstrates auto-detection of phases, long-format reshaping, and two
analysis tiers: (1) 2-D per-FOV quantities — phase fraction, mean particle
size, and interparticle spacing — summarised and plotted in Nature journal
style; (2) 3-D stereological quantities — volume fraction $V_V$, surface
area density $S_V$, mean caliper diameter $\bar{D}$, and 3-D mean free path
$\lambda_{3D}$ — derived using stamp.stereo.volume_fraction,
surface_area_density, mean_caliper_diameter, and mean_free_path_3d,
then summarised and plotted in the same style. All tables exported as CSV
and LaTeX booktabs via stamp.export.