API Reference

`thermosim.utils.io`

`load_irb_txt`

load_irb_txt(path: str | Path) -> tuple[np.ndarray, dict, dict]

Load an IRB-style CSV thermogram exported from the IR camera software.

Parameters

Name	Type	Description
`path`	`str` or `Path`	Path to the `.csv` / `.irb` text file

Returns

Name	Shape	Description
`data`	`(H, W)` float32	Temperature array in °C
`settings`	`dict`	`[Settings]` section (image size, calibration range, …)
`params`	`dict`	`[Parameter]` section (filename, frame index, timestamp)

Example

from thermosim.utils.io import load_irb_txt

arr, settings, params = load_irb_txt("frame_0001.csv")
print(arr.shape)        # (333, 507)
print(settings["TempUnit"])   # '°C'
print(params["RecTime"])      # '14:41:36'

`thermosim.pore_generator`

`PoreGenerator`

Places non-overlapping spherical pores randomly inside the unit cube \([0,1]^3\) and produces a FEniCSx-compatible diffusivity field.

PoreGenerator(
    n_pores:   int   = 20,
    r_min:     float = 0.03,
    r_max:     float = 0.08,
    margin:    float = 0.005,
    max_tries: int   = 5000,
    seed:      int   = 0,
)

Methods

`is_pore(x)`

is_pore(x: np.ndarray) -> np.ndarray

Vectorised membership test. x has shape (3, N) (FEniCSx coordinate layout).
Returns boolean array of shape (N,).

`make_diffusivity_field(msh, alpha_matrix, alpha_pore)`

Returns a DG0 FEniCSx Function with alpha_matrix outside pores and alpha_pore inside.

gen = PoreGenerator(n_pores=20, r_min=0.03, r_max=0.08, seed=42)
alpha = gen.make_diffusivity_field(mesh, alpha_matrix=1e-6, alpha_pore=2e-5)

`summary()`

Prints number of placed pores, porosity (%), and radius range.

`scripts/calibrate.py`

Command-line script — not imported as a module.

python scripts/calibrate.py <input.h5> [--output <calib.npz>]

Argument	Description
`input`	HDF5 file with dataset `thermogram` of shape `(N, H, W)`
`--output`	Output `.npz` path (default: `<input>.calib.npz`)