菜单
Create a field desorption map (detector hit-density / concentration image) and optionally a movie of detector activity over the course of the experiment. Use when the user wants an FDM, desorption map, detector image, or to watch features evolve during the measurement.
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
基于 SOC 职业分类
Voronoi-cell cluster determination (Felfer et al. 2015) — detect and extract clusters via Voronoi-cell volumes, Delaunay triangulation, and a Kolmogorov-Smirnov test against a randomised reference. Use when the user wants the Voronoi/clusterDetermination method specifically (this is NOT DBSCAN — for density-based DBSCAN use /apt-cluster instead).
Create a new ion color scheme or add ions/colors to the existing one. Produces/updates the `colorScheme` workspace variable used by ranging and 3D visualization everywhere else. Use when the user wants custom ion colors, a new color scheme, or to add a missing ion color.
Per-vertex concentration mapping onto a mesh/interface — compute atom-to-mesh distances, clip a slab around the interface, voxelise by nearest vertex, and apply a concentration kernel to colour each vertex. Use when the user wants a concentration map painted onto an isosurface or modelled grain boundary mesh.
Run quality-control checks on an atom probe dataset before analysis — basic stats, voltage curve, detector hit map (FDM), mass spectrum sanity, coordinate ranges, multi-hit fraction, and automated quality metrics. Use when the user wants to assess data quality or sanity-check a reconstruction.
Quantify interfacial excess (Gibbsian excess of solute, in at/nm^2) across an interface — either a single value via the Krakauer-Seidman integral method, or a 2D excess map over the surface. Use when the user wants the amount of an element segregated to an interface or grain boundary.
Full pipeline from voxelisation through proxigram — voxelise, build an isoconcentration surface, then compute single- and multi-ion proxigrams across it in one guided sequence. Use when the user wants a proxigram but does not yet have an interface.
| name | apt-fdm |
| description | Create a field desorption map (detector hit-density / concentration image) and optionally a movie of detector activity over the course of the experiment. Use when the user wants an FDM, desorption map, detector image, or to watch features evolve during the measurement. |
| argument-hint | [species] |
Create a field desorption map (FDM) from the detector hit coordinates, and optionally a movie of how the detector pattern evolves during the experiment. Use the MATLAB MCP server.
The MATLAB workspace must contain:
pos — atom probe data with detector coordinates detx, dety (in nm). These exist only in extended formats (.epos / .apt / pyccapt HDF5), not plain .pos. If only a .pos is loaded, tell the user to reload via /apt-load with an .epos/.apt/HDF5 file.pos must also be allocated (have an ion/atom column) — run /apt-composition first so a spec with ranges exists, then allocate. A plain hit-density FDM needs no ranging.dist = [pos.detx, pos.dety];
bin = [0.07 0.07]; % bin width in nm for x and y -> FDM resolution
mode = 'distance'; % only valid 2D/3D mode
[binCenters, binEdges] = binVectorsFromDistance(dist, bin, mode);
clearvars bin mode
Smaller bin = finer FDM but more pixels. 0.07 nm is the workflow default.
Fastest option — a 2D histogram of all detector hits. from/to select an ion-index window, which corresponds to a slice of the experiment in time (lower indices = earlier in the run):
from = 1;
to = height(pos); % use all hits; narrow this to look at a time window
FDM = hist3([pos.detx(from:to), pos.dety(from:to)], binCenters);
figure; imagesc(FDM); axis equal;
clearvars from to
Voxelise the detector plane and compute the local concentration in each pixel, then display one species. $ARGUMENTS is the species to map (e.g. Al):
startIdx = 1;
endIdx = height(pos);
pix = posNdBin(pos(startIdx:endIdx,:), dist(startIdx:endIdx,:), binEdges);
detEff = 82; % detector efficiency (% or decimal both accepted)
excludeListConc = {'unranged', 'Ga'}; % species to ignore (optional)
volumeName = ''; % volume name, optional
concentrationKernel = @(p) posCalculateConcentrationSimple(p, detEff, excludeListConc, volumeName);
conc = cellfun(concentrationKernel, pix, 'UniformOutput', false);
wantedSpecies = cellfun(@(p) p.Al(p.format == 'concentration'), conc); % replace Al with $ARGUMENTS
figure;
imagesc(binCenters{1}, binCenters{2}, wantedSpecies);
axis equal;
Confirm the detection efficiency with the user (depends on instrument) and which species column to extract. Bin size strongly affects compute time here.
Stack frames sampled across the run into an .avi. Uses all species (hit density). Plays at 30 fps:
frames = 300; % number of frames (movie length = frames/30 s)
sample = 10000; % ion hits per frame
RES = 100; % FDM resolution RES x RES (>1)
fileName = "FDM"; % output .avi saved in current folder
movieFDM = movieCreateFieldDesorptionMap(pos.detx, pos.dety, frames, sample, RES, fileName);
frames, sample, RES, fileName are all optional in the function. The movie reveals grain boundaries, poles, and phases "moving" as the tip is consumed.
Spatial or temporal cropping is done by hand:
from/to ion-index window (step 2a) or the frame sampling (step 3) to focus on a part of the run.plotFieldDesorptionMap(pos) opens a figure with an interactive circular spatial crop. Tell the user to drag the circle in the MATLAB figure to keep only the hits inside it — do not try to automate the mouse interaction.Tell the user which FDM was produced (plain hit-density vs. species concentration), the bin size/resolution used, and where any movie file was saved. Note that the FDM axis units are arbitrary for the hist3 version and scale with bin width. Suggest:
bin/RES for more or less detail.from/to).plotFieldDesorptionMap(pos) for interactive inspection and cropping.binVectorsFromDistance, posNdBin, posCalculateConcentrationSimple, movieCreateFieldDesorptionMap, plotFieldDesorptionMap (optional, interactive); hist3, cellfun, imagesc (MATLAB built-ins).