Alligator - Prediction and Validation Input
Prediction and Validation Input
To predict a sample set with a calibration model please
- select a model or a model set on the left
- select a sample set on the right
Press "Predict" or "Validate".
Please select only one sample set and only one model set. If you select PLS, LPS, LOCAL, SVM, ANN or PCA models you may select a´multiple models INSIDE each one kind of model. I.e. you cannot select a single PLS and a single MLR model to be predicted simultaneously.
If you want to predict multiple sample sets and compare the results please go to "Network" in project view to select "network predictions".
If you have activated grouding in the project settings you can select "Use all models in group" to select only one model but predict with all models in that group of models. Alternatively you can mark all models in a group.
if the target numbers of the sample set and the model(s) / model set mismatch a popup dialog will present appropriate transfer models in this project.
An appropriate transfer model is defined by matching source number for transfer model and spectra and matching target number for transfer model and calibration model to apply:
sample set | transfer model | model | |||
target number: | 123456 | source number: | 123456 | ||
target number: | 654321 | target number: | 654321 |
Validation results may be saved to the project and retrieved later on the right side of the input dialog.