Changelog

All notable changes to this project will be documented in this file.

The format is based on Keep a Changelog, and this project adheres to Semantic Versioning.

[0.4.0] - 2025-10-21

A button to visualize the selected heat range, i.e. add vertical markers in the plot.
A "Save Analysis" button to add the current analysis to the downloadable Excel file
The analysis now includes a plot for the metabolic load itself.
The downloadable Excel file with the analysis results now also contains the data which can be used to plot the curves shown in the UI, i.e. the original measurement data in the time and enthalpy domain, as well as the fitted curves.
The parameter rₗ, i.e. the growth rate on the logistic curve, is now reported.
METABOLATOR now reports its own version based on the Git reference it is using.
The METABOLATOR web application can now be started using the command metabolator-web.

Simplified the fitting routine. It used to narrow the fitting region to find the best possible fit. Now, the selected time range is fitted in its entirety. This makes the procedure much faster.
Simplified the four point method used to determine the starting values for the fit.
If the four point method fails, a fallback method is used to find starting values.
The parameter r₀ was renamed to rₘ to signify the growth rate of the Monod equation.
Growth rates are now reported in 1/h, rather than 1/s.
The fitting module is now based on lmfit's Model class.
When plotting in the time domain, the time is now shown in hours rather than seconds.
The downloadable Excel file only contains analyses that were previously added by pressing the new "Save Analysis" button, rather than all ran analyses.
The labels on buttons and in the plot are now more intuitive and distinguish better between measured data and fitting results.
The "Manual" link in the header now links to a page on the HZDR website rather than downloading a PDF.
The dependency on Poetry was removed. METABOLATOR can now be used and developed with only a Python installation (Python, venv, Pip).

The fitting results are now reported on the selected interval on which they were actually calculated.
Negative measured values at the start of the curve are now excluded and enthalpy is calculated only after excluding the negative values. This prevents the occurrence of negative values in the enthalpy plot.
An off-by-one error in the curve_start method was fixed.

Datasets can now be loaded from DOI/URL which is implemented in the remote_data module.
The software version and a "Contribute" link are now shown in the footer.
Search links to find datasets on Rodare and Zenodo are now shown in the header text.

The UI layout was changed to accomodate for the new remote data functionality. In the course of this change, large parts of the ThermalAnalysisApp class were refactored.

Large update to the UI, moving away from the grid layout.
The function plot_fitting in metabolator.plots was renamed to plot_fitting_enthalpy.
The function plot_fitting_time in metabolator.plots now takes an additional title parameter.

Non-optimal fitting intervals are now caught and don't cause the app to get stuck any more.
If the fitting procedure fails due to bad selected range or offsets, this is now correctly reported back to the user and doesn't make the app get stuck any more.
Actually use the ranges determined by the four point method in the fitting process. Previously, starting values from the four point method were used but the range was applied and then overwritten.
The Pmax of a fit is now correctly reported on the fitted curve rather than the measured curve.

Initial release of the METABOLATOR analysis tool.