X Ray Diffraction - Theoretical Absorption Factors

A GUI app to compute theoretical xray absorption factors (for xray diffracted in a capillary) efficiently on the device's GPU. Made with C#/Unity.

AcediaKleea	d8xa	zzfab

Download the latest release here. Extract the zip file and put the folder wherever you want to run the app.

The app uses presets to store all configurable parameters of the experiment setup. The main screen has input fields for all those parameters.
To save or overwrite a preset, enter a name into the field preset name and hit the save button.
To load an existing preset, enter the name and click load.

All presets are stored under XRD-TAF/Presets, in .json format and can be copied to other installations of the app.

There are three experiment modes available:

• Point: Computes a single averaged absorption factor for a single path of the xray (diffracted horizontally by angle 2theta).
• Area: Computes pointwise averaged absorption factors for all xray paths to the points on a detector screen.
• Integrated: For each horizontal diffraction angle 2theta, a ring segment is projected onto a detector screen, starting from the intersection point of xray and detector. A sequence of k points is distributed uniformly over that ring segment. The averaged absorption factors for each point are then combined into an average over the whole ring segment.

The GUI only displays the parameters relevant to the currently selected mode.
Here's an overview of all parameters:

Contains global settings such as flags and default values. All settings have a description text.

Inside the XRD-TA folder containing the app, this is the folder/file structure:

├── Benchmark 
│   ├── benchmark.csv
│   └── config.csv
├── Input
│   ├── angle-list1.txt
│   ├── angle-list2.txt
│   ├── angle-list3.txt
│   └── ...
├── Logs
│   └── mode{mode}_debug.txt
├── Output
│   ├── group
│   │   ├── preset1
│   │   │   ├── [mode=<mode>][dim=(<res>,<n>,<m>,<k>)] Output.txt
│   │   │   └── ...
│   ├── preset2
│   │   └── [mode=<mode>][dim=(<res>,<n>,<m>,<k>)] Output.txt
│   ├── preset3
│   │   └── [mode=<mode>][dim=(<res>,<n>,<m>,<k>)] Output.txt
│   └── ...
├── Presets
│   ├── preset1.json
│   ├── preset2.json
│   ├── preset3.json
│   └── ...
└── Settings
    └── settings.json

All output directories will be created when needed, if not present.

Stores benchmark config and results.

config.csv is a user-generated table containing parameters for each iteration of the benchmark.
<timestamp> benchmark.csv contains a table of benchmark results for config.csv, generated at the timestamp.

Benchmark parameters are as follows:

mode is a value in (0,1,2), which stand for 0=point, 1=area, 2=integrated.
res is the grid resolution of the capillary cross-section.
n: The number of angles to use in the benchmark. In mode 1, n this is equivalent to the number of pixels on the horizontal axis of the detector.
m: The number of pixels on the vertical axis of the detector. (Ignored in mode 0 and 2).
k: Amount of angles in the projected ring segment of mode 2. (Ignored in mode 0 and 1).

The directory where the app searches for angle lists. Lists are expected to have one value per line and no header.

If enabled in the settings, log files will be saved in this folder. The log filename will be <timestamp> log.txt.

Calculated absorption factors will be saved in this folder. They will be grouped in folders by preset name -- and if a parent folder is specified in the preset, the whole preset folder will be put in a parent folder see section [Folder structure](### Folder structure).

The filename for each output is of the format [mode=<mode>][dim=(<res>,<n>,<m>,<k>)] Output.txt.

The directory where are stored, in the format <preset name>.json. The name specified in the preset is identical to the filename (without extension).

Stores global user settings in settings.json.

The pre-defined preset benchmark.json will be used for benchmarks. Changes to this preset do not influence the benchmark, since the app takes all time-complexity-relevant parameters from config.csv. Angle files will be ignored, since angle lists will be generated from the parameter n, i.e. n values evenly distributed between 0 and 180°.

To add a benchmark iteration to the config, add a line containing the parameters mode, res, n, m k in the exact same order to Benchmark/config.csv.

Example:

Parameter m will be ignored by modes 0 and 2, k will be ignored by mode 0 and 1.

The app expects tabs as value separators for the csv.

All float input values, in the GUI and in input files, are expectd in invariant culture style, i.e. with . as decimal separator.

All .csv files have to use tabs as separators.