Results¶

GET /analysis-jobs/{job_id}/result/ returns a completed job's outputs:

{
  "job_id": "…",
  "metrics": {
    "solution": { "…": "…" }
  }
}

Everything scientific is in metrics.solution. Its keys depend on the analysis. This page lists them all, with the headline the app shows for each and what's common to every result. The Scientific reference defines the quantities.

Common to every solution¶

Values are converted to the image's current voxel size every time you fetch the result. See Units and voxel size.

Example: pore size distribution¶

A real result, for the 120³ synthetic sample used in the figures on these pages, with a voxel size of 5.345 µm (tables trimmed):

{
  "job_id": "1b7e5c2a-8d4f-4e1b-9a3c-6f2d8e0b4a7c",
  "metrics": {
    "solution": {
      "R": [5.345, 10.69, 16.035, 21.38, 26.725, 32.07, 37.415, 42.76],
      "pdf": [0.027333, 0.014392, 0.041666, 0.044111, 0.043736, 0.011283, 0.004251, 0.000319],
      "cdf": [1.0, 0.85391, 0.77698, 0.55428, 0.3185, 0.084736, 0.024426, 0.0017038],
      "satn": [0.14609, 0.076928, 0.2227, 0.23577, 0.23377, 0.06031, 0.022722, 0.0017038],
      "bin_centers": [5.345, 10.69, 16.035, 21.38, 26.725, 32.07, 37.415, 42.76],
      "bin_edges": [2.6725, 8.0175, 13.362, 18.707, 24.052, 29.397, 34.742, 40.087, 45.432],
      "bin_widths": [5.345, 5.345, 5.345, 5.345, 5.345, 5.345, 5.345, 5.345],
      "sizes": 8,
      "steps": 25,
      "voxel_size": 5.345,
      "voxel_size_um": 5.345,
      "units": {"length": "µm", "area": "µm²", "volume": "µm³", "pressure": "Pa"},
      "tables": [{"title": "Pore Size Distribution", "columns": ["R (µm)", "pdf", "cdf", "…"], "rows": ["…"]}]
    }
  }
}

The median pore radius, where cdf crosses 0.5, is about 22.6 µm.

Keys by analysis¶

Lengths are in µm, areas in µm², volumes in µm³ and pressures in Pa once the image has a voxel size, and in voxels, voxel², voxel³ and N/m per voxel otherwise. The headline is the one number the app, job lists and reports show first; where it isn't a key of its own, its line says how to read it from the keys.

permeability¶

Headline: k_mD, or k_voxel2 when the image has no voxel size.

Palabos returns the permeability keys, converged, iterations_completed and max_iterations only.

Some solvers also name the permeability permeability [m^2] and permeability [mD]. They're the same values as k_m2 and k_mD, and like them follow the image's current voxel size.

diffusivity¶

Headline: tortuosity, the tortuosity factor \(\tau\).

Porosity isn't returned separately, but \(\varphi = \tau \cdot D_\text{eff}/D_0\) recovers it, and ingestion records it on the image. The formation factor \(F = D_0/D_\text{eff}\) isn't a key of its own either: it's 1 / effective_diffusivity, and summary_tables carries it as the row Formation factor (F = D_0/D_eff), left out when effective_diffusivity is 0.

conductivity¶

Headline: effective_conductivity, the relative conductivity \(\sigma_\text{eff}/\sigma_s\).

Plus the convergence keys of diffusivity. Everything is dimensionless, so the voxel size doesn't affect the result, and there's no formation factor row.

poresize¶

Headline: the median pore radius \(r_{50}\), where cdf crosses 0.5, read against R.

hg_porosimetry¶

Headline: \(P_{c50}\), where snwp_intrusion crosses 0.5, read against pc_intrusion.

The curves are padded at their ends to reach full saturation (intrusion) and the trapped saturation (extrusion). The padded end pressures are infinite and appear as null.

radial_density¶

Headline: \(r_{50}\), the median distance to solid (to pore, for the solid phase), where cdf crosses 0.5, read against R.

chord_length_distribution¶

Headline: the median chord length \(\ell_{50}\) along x, where distributions.x.cdf crosses 0.5, read against distributions.x.L.

phase (the phase measured), axis, bins and ndim, and distributions.x, distributions.y and distributions.z, each with:

An axis with no complete chords, for example because the image is too thin along it, comes back with empty lists.

porosity_profile¶

Headline: the mean porosity \(\bar\varphi\), the mean of every value in profiles.{x,y,z}.profile.

axis, ndim, and profiles.x, profiles.y, profiles.z (whichever were requested), each with coords (slice midpoint positions, length) and profile (slice porosity).

rev_porosity¶

Headline: the porosity of the largest regions, the mean of porosity over the 10% of cubes with the largest volume.

sem_view, xray_view¶

Headline: none. sem_fields_written or xray_fields_written lists the fields written to the image.

Fields¶

Some analyses also save 3D fields with the image. They're shown in the app's slice viewer and can be downloaded from the app's Export, but downloads aren't in the API yet: it returns a result's numbers only. Deleting the image deletes its fields; the numbers in solution stay.