In routine analysis samples need to be evaluated using previously designed
calibration models. This procedure is called prediction. The result is
either a qualitative information, which identifies a product or compound
or quantitative, which produces a concentration or similar value. The
software provides several opportunities to present prediction results
as described in more detail in the following:
Some applications simply require a quick overview of the prediction
results without the need for all the statistical details of a report.
Displaying the predicted value, e.g. the predicted concentration is sufficient
in most cases. The auto-evaluation
tool of the software provides an online prediction capability, which
shows the result for the active data object on screen immediately.
The name of the calibration and the predicted value is shown in the
top left corner of the data view as illustrated in the screenshot below:
If more than one data objects are displayed merged in a single data
view, the active object is shown emphasized. The results are updated automatically
for the current active data object in the data view.
A more comprehensive prediction result is provided in a detailed evaluation
report. Herein all statistical results of the calibration plus the prediction
results are listed together with a spectrum screenshot in one report.
The report can be easily printed out from the software using the Print
command. Alternatively it can be copied into the clipboard to be pasted
into other office applications. All the details provided in the reports
satisfy most requirements to give evidence for CFR21 part 11 regulated
environments. A sample report looks like this:
A report can be easily created for a single or multiple calibrations
Creating a prediction report for a single calibration
Please refer to the "Evaluate
with..." command section.
Creating a prediction report for multiple calibrations
Please refer to the "Evaluate"
Herein all calibrations of all open projects are considered. However,
only those calibrations are considered, which contain suitable evaluations
for the actual data object.