User flows

⚙️ Generated page — do not edit here. Source: docs/user-flows_en.md. Edit it in the repository and the wiki rebuilds itself.

User flows

This guide summarizes the usual journeys in Fitbauer's Qt interface.

1. Open a spectrum and simulate

1. Start the application:

   python fitbauer.py

2. Open a file from File → Load... (.ws5 or .adt).

3. Review the folding point and Vmax in the calibration panel.

4. Adjust component parameters manually if desired.

5. The plot updates in simulation mode until a real fit is launched.

2. Discrete fitting

1. Choose discrete mode.
2. Select the number of components.
3. For each component:
   • type: singlet, doublet or sextet,
   • initial values,
   • fixed/free parameters.
4. Optional: use Initialize from minima to propose components.
5. Press Fit.
6. Review:
   • χ², reduced χ², AIC/BIC,
   • residual diagnostics,
   • percentage areas,
   • errors and correlations if available.

3. Bootstrap

1. Perform or prepare a discrete fit.
2. Go to Tools → Bootstrap.
3. Choose the number of replicas.
4. The result updates the uncertainties and the error source indicated in reports.

4. Distribution P(BHF) / P(ΔEQ)

1. Switch the mode to distribution.
2. Choose the variable: P(BHF) or P(ΔEQ).
3. Configure:
   • minimum/maximum range,
   • number of bins,
   • δ, fixed ΔEQ/BHF, Γ,
   • log10 α, shape and regularization.
4. Optional: use L-curve α to estimate regularization.
5. Optional: enable sharp components to add discrete phases to the fit.
6. Press Fit.
7. Review the fitted curve, residual and resulting distribution.

5. Calibration

There are two routes:

• Web calibration: download metadata from the laboratory API.
• Local calibration: use the current file as calibration from the context menu.

The calibration affects traceability, Vmax and the isomer shift reference when available.

6. Save session

Use File → Save session... to save:

• file/counts,
• model,
• parameters and fixed flags,
• fitting options,
• distribution,
• relevant visual preferences.

The JSON session keeps compatibility with historical formats.

7. Export results

Main options:

• Save fit: exports velocity, data, model and residual in text.
• Export Plotly: interactive HTML figure.
• Export report: Markdown and, if available, PDF.

The report includes traceability, calibration, parameters, areas, metrics, correlations and residual diagnostics.

8. Validations and warnings

Before fitting, Fitbauer validates:

• data lengths and finiteness,
• parameter bounds,
• positive sigma,
• distribution ranges,
• minimum number of bins,
• known component types and regularization.

If there are problems, the GUI shows a warning and does not launch the fit.