What this workflow helps you do
Compare two or more experimental datasets in one NotyPlot workspace without losing track of units, sample identity or display transformations. Overlay, stack or normalize series only after confirming that the datasets are scientifically comparable.
Use this workflow when
- Comparing samples, treatments, batches or repeated experiments.
- Showing peak or trend shifts across datasets while keeping sample identity clear.
- Preparing an overlay or stacked comparison for a report or publication.
What you need before you start
- Two or more datasets whose variables and units can be aligned meaningfully.
- Clear sample names and acquisition context.
- A decision about whether absolute magnitude, normalized shape or vertically separated curves best answer the comparison question.
Step-by-step workflow
- Import each dataset and give every series an unambiguous sample name.
- Verify that the X variables, units and relevant acquisition conditions are compatible.
- Choose overlay, stacked or normalized comparison according to what must remain visible.
- If normalization or vertical offsets are used, inspect how they change the apparent differences between samples.
- Set a readable legend and line presentation; avoid styles that make one sample difficult to distinguish.
- Export the comparison and document any normalization, offset or preprocessing that affects interpretation.
What to review in the result
- Datasets use compatible units and comparable axes.
- Normalization is used for shape comparison rather than mistaken for an absolute-intensity comparison.
- Vertical offsets are presentation choices and are not interpreted as physical changes.
- Every trace is identifiable in the legend and caption.
Before you report or export
- State when curves were normalized, stacked or offset.
- Keep raw values available for any comparison that depends on absolute magnitude.
- If collection conditions differ, describe those differences rather than treating the traces as directly equivalent.
Limitations and interpretation
- Visual overlap does not prove statistical or mechanistic equivalence.
- Normalization can suppress real amplitude differences, while stacking can exaggerate apparent separation.
- Datasets with different units or fundamentally different acquisition conditions may require transformation or separate interpretation before comparison.