Auto-plot suite
A full plot suite, chosen and drawn automatically.
Niobia chooses the right plot for the data and draws it to the convention the method demands: equal-aspect Nyquist plots, frequency-colored Bode plots, dQ/dV through a resample-smooth-differentiate pipeline, capacity fade with end-of-life lines. A whole figure suite, with no manual charting.
What it measures
A plot is not neutral. Each electrochemical and materials method has a convention that makes its information readable, and breaking the convention hides the signal:
- Nyquist must be equal-aspect, or the semicircle that carries the charge-transfer resistance is distorted into an ellipse and misread.
- Bode wants magnitude and phase against a log-frequency axis, and frequency coloring so the reader can locate the time constants.
- dQ/dV cannot be drawn from raw cycler data; it needs a resample, then smooth, then differentiate pipeline, or noise becomes phantom peaks.
- Capacity fade, rate capability, Ragone, Tafel each have an expected axis pair, scale, and reference line (the 80% end-of-life line, the benchmark current density) that turn a curve into a decision.
The suite is the set: for one dataset, the full collection of figures a competent analyst would draw, chosen by what the data is rather than by a template the user filled in.
How to read the output
Judge an auto-generated figure the way you would judge a colleague's: is the axis pair right for the method, is the scale (linear versus log) correct, is the convention honored, and is the reference line present. A correct Nyquist is square; a correct Bode is log-frequency; a correct dQ/dV is smooth without being scrubbed of its real peaks. The point of automating the suite is not to skip judgment but to remove the manual charting labor and the convention errors that creep in when every plot is rebuilt by hand under deadline.
A real use case
An engineer finishes a rate-capability and EIS study on a new electrolyte and needs the figure set for a design review tomorrow. By hand, that is an afternoon of charting: getting the Nyquist aspect ratio right, coloring the Bode by frequency, building the capacity-versus-C-rate plot and the Ragone, and redrawing two of them after a column mix-up. From the ingested data, the suite comes back drawn to convention in one pass: equal-aspect Nyquist, frequency-colored Bode, polarization-overlaid rate plot, Ragone. The engineer spends the evening reading the result and writing the conclusion instead of fighting plotting defaults.
Common mistakes
- Drawing a Nyquist plot without equal aspect. The distorted semicircle leads to a wrong charge-transfer resistance read by eye.
- Differentiating raw data for dQ/dV. Skipping the resample-and-smooth step turns measurement noise into peaks that get interpreted as chemistry.
- Linear axes where the physics is logarithmic (impedance frequency, some kinetics), flattening the feature you needed to see.
- Omitting reference lines: a fade plot without the 80% line, a Tafel plot without the benchmark current density, leaves the reader to estimate the decision point.
- Treating auto-plotting as a reason to stop checking. The convention is automated; the interpretation is still yours.
Method-aware plotting, not default axes
Niobia picks the plot type from what the data is and renders it to the method's convention: equal-aspect frequency-colored Nyquist plots, Bode magnitude and phase against log frequency, dQ/dV and dV/dQ through the resample, smooth, differentiate pipeline that avoids differentiation artifacts, capacity-fade curves with the end-of-life line, rate-capability and Ragone plots, Tafel and Levich constructions. Because it draws the whole suite a method warrants rather than a single templated chart, the output is the figure set a review needs, built from the ingested data and ready to drop into a report. The drawing is automated; the conventions are the ones a publication would demand.
Frequently asked
Why does plot convention matter so much?
Because conventions are what make a method's signal legible. An equal-aspect Nyquist, a log-frequency Bode, a properly smoothed dQ/dV: break any of these and the chart still renders but the feature you needed (a resistance, a time constant, a degradation peak) is distorted or hidden.
Does auto-plotting replace the analyst?
No. It removes the manual charting labor and the convention errors, and produces the figure set a competent analyst would draw. Interpreting the figures, deciding what they mean for the design, remains the engineer's job.
What does a plot suite include for a typical battery study?
Whatever the data supports: capacity fade with the 80% line, voltage profiles, dQ/dV overlays, Nyquist and Bode for impedance, rate-capability and Ragone plots, and the relevant reference lines. The set is chosen by the data, not assembled from a fixed template.