PEAXACT 6 (R2025) Release Notes
Version 6 (R2025): 2025-04-24
- Samples can now be given a display name using the new special feature
{Title}. Titles can be read from certain file types (JCAMP, Magritek, Bruker TopSpin, JEOL), or can be auto-created from the file name or path. - Support for Jeol Delta files (.jdf)
- Support for Bruker TopSpin files (1r, 2rr).
- Support for Viavi files (.sam).
- Support for reading exact timestamps from LabCognition/panorama SPC-files.
- (Model Tree) Supports mouse drag & drop to reorder model components.
- (Model Tree) Supports mouse drag & drop to copy components between models.
- (Pretreatment Model) New alignment methods based on pattern matching.
- (Pretreatment Model) New smoothing method (replaces old method) with better exactness for sharp peaks. Old models using the older method are still supported, but for new models the old option is not available anymore.
- (Pretreatment Model) New resampling method
Resolution. - (Pretreatment Model) New standardization method
Variance norm. - (Pretreatment Model) Standardization methods
MinMax normandArea normhave been removed. Old models using these methods are still supported, but for new models the old options are not available anymore. - (Hard Model) Mouse-modification of component weights (can be enabled by new user preference).
- (HMFA) New algorithm which now also works for samples with different x-axes and uses the user-defined baseline instead of a linear baseline.
- (Classification Model) New classification methods: Nearest Neighbors (NN), PCA Quadratic Discriminant Analysis (PCA-QDA), and Partial Least Squares Discriminant Analysis (PLS-DA).
- (Classification Model) Classification method
Database lookuphas been replaced in favor ofNearest Neighbors. Old models using the old method are still supported, but for new models the old option is not available anymore. - (Classification Model) New diagnostic
Class Similarity. - (Classification Model) New option to enable identified groups for a subsequent groupwise prediction. Note: This is the reason why the Classification Model is now displayed before the Calibration Model in the Models Panel.
- (Classification Setup) Removed CV option
leave-1-outas it is not recommended for classification. - (Calibration Model) Groupwise calibration now supports choosing individual regression types per group.
- (Prediction) Changed behavior for groupwise predictions, i.e., predictions made with a groupwise Calibration Model. Before, the predicted value for an invalid/uncalibrated group was calculated using a fallback calibration. However, it wasn't always obvious which calibration was used. Now, if a prediction is made for a group that isn't calibrated, the value will be missing (NaN). This change makes it clearer as predictions are only made using the groupwise calibration, and if there's no proper calibration for a group, no prediction will be made at all.
- (Custom Model) Custom results can now be marked as intermediate results, hiding them in all reports.
- (Custom Model) Added new functions for custom expressions:
EXP,LOG,ABS,LOWER,UPPER,FEATURE,ISFEATURE. - (Custom Model) Function
IF()now supports more than one condition as a better alternative to nestedIF-statements. - (Custom Model) Results with broken references appear crossed out in the Models Panel.
- (Custom Model) Auto-repair of broken references when replacing a Calibration Model or Classification Model.
- (Report Generators) Quick access to styling options for reports.
- (Report Generators) New styling options for plot lines, plot markers, grid lines.
- (Report Generators) Better formatting of timestamp labels in graphical reports.
- (Report Generators) Consistent use of markers throughout all reports.
- (Calibration Report Window) Coloring by feature.
- (Validation Report Window) Coloring by feature.
- (Data Inspector) Assign feature values to selected samples in the table.
- (Data Inspector) Ranking samples for calibration.
- (Data Inspector) Added OR-filter and AND-filter to create complex filter expressions for the data table, e.g., multiple choice filters and range filters.
- (Data Inspector) New graphical report
Samples (Stack). - (Data Inspector) Assign feature values to selected samples in plots.
- (Data Inspector) The Table Editor now highlights selected rows and supports selection of whole rows by clicking on the row number.
- (Data Inspector) The Join Table Dialog now visually highlights affected rows.
- (Command Line Interface) New command line option
-analyzeto perform any analysis from the command line. - (Command Line Interface) New command line option
-resultfile. - (License Activation) New Update-button to quickly update the license when it changed (renewal, new version, etc.).
- Several bug fixes and stability improvements.
- (AppServer) NET API is now available for NET 6.0+ without dependencies to .NET Framework.
- (AppServer) Removed version-independent ProgId.
- (AppServer) COM API supports creating samples with complex y-values.
- (AppServer) New and improved diagnosis utility.
- (AppServer)
PerformAnalysis()withAnalysisType.Autonow returns prediction results instead of identification results if a model has both a Calibration Model and Classification Model (but no Custom Model). - (AppServer) NET API namespace renamed to
S_PACT.Peaxact.AppServer. The old namespace (S_PACT.PEAXACT) is still supported for backward compatibility.
- ProcessLink 6 OPC UA server uses a new security certificate. Server and clients must exchange certificates again.
- ProcessLink 6 OPC UA server uses a new TCP port (61101).
Compatibility
- PEAXACT 6 is backward compatible. You can load model and session files saved with earlier versions. Model and session files saved with PEAXACT 6 cannot be loaded by earlier versions though. Please review all upgrade notes before you upgrade.