What’s New in Release 2025?
- New classification methods
- Synergy of classification and calibration
- Color-coding calibration samples
- New stack plot
- Improved plot styling
- Sample ranking for calibration
- Improved Pretreatment Model
- Improved Custom Model
- Advanced data table filtering
- {Title} for samples
- Quality of life improvements
- View all release notes
New Classification Methods
PEAXACT 6 (R2025) introduces three new classification methods: Nearest Neighbors (NN) - a better replacement of the Database Lookup method - and two discriminant analysis methods which are based on principal components (PCA-QDA and PLS-DA).
Synergy of Classification and Calibration
The Classification Model and Calibration Model now work closer together. Identified groups from a Classification Model can be used for a subsequent prediction, using a groupwise Calibration Model. Groupwise calibration now supports individual regression types per group. Use the combination, for example, to identify a sample’s product category first, then predict a product-specific composition – all within a single model.
Color-coding Calibration Samples
PEAXACT 6 (R2025) enables you to investigate the influence of (uncalibrated) features by visually highlighting calibration results with distinct color codes.
New Stack Plot
The Data Inspector features a new stack plot, enabling vertical stacking of spectra with optional grouping and Profile Cursor support.
Improved Plot Styling
Graphical reports offer new styling options for plot lines, markers, and grid lines. Styling features are easily accessible from the toolbar.
Sample Ranking for Calibration
Samples can be ranked by their importance for calibration based on the diversity of their features. This simplifies the selection of a small, uniformly distributed subset from a larger set of often repetitive, real-life samples and offers a practical alternative to traditional design-of-experiments methods when synthetic calibration samples aren’t feasible.
Improved Pretreatment Model
PEAXACT 6 (R2025) adds new pretreatment options: alignment can now be performed using pattern matching, resampling is available with a fixed resolution, and the new standardization option "Variance norm" makes it easier to compare spectra of different scales.
Improved Custom Model
Custom results can now be marked as intermediate, allowing you to hide them from all reports. In addition, new and enhanced functions for custom expressions are available, such as IF() with multiple conditions and FEATURE() for including sample-specific feature values.
Advanced Data Table Filtering
The Data Inspector supports complex filter expressions for the data table, including multiple choice filters and range-based filters for more flexible data selection.
Titles for Samples
Samples can be assigned a display name, allowing you to control how samples are labeled in visualizations, using the new special feature {Title}. Titles can be imported from supported file types (JCAMP, Magritek, Bruker TopSpin, JEOL) or automatically generated from the file name or path.
Quality of Life Improvements
PEAXACT 6 (R2025) introduces several small but impactful improvements to make your daily work easier. For example, you can now reorder model components or copy them between models with simple drag & drop. Additionally, feature values can be directly assigned to selected samples in Data Inspector plots, without needing to switch to the table. See the release notes for more.