A measurement aid for points you click. Not laboratory instrumentation, not regulatory evidence, and not a substitute for source data. The area comes from trapezoids between points you placed by eye on an image, scaled by a two-point calibration you also placed by eye. Calculations assume linear X and Y axes. Use at your own risk. Open for the full scope limits.
What this tool actually does
It is a visual measurement and numerical integration aid. It converts pixel positions to data coordinates using the calibration you set, then integrates the points you placed by the trapezoidal rule. It is not professional engineering advice, regulatory evidence, laboratory instrumentation, or a substitute for the source data.
Linear axes only
Calculations assume linear X and Y axes. Do not use this tool for logarithmic, categorical, distorted, perspective or transformed plots unless you have independently converted and validated the data first. A log axis read as linear does not produce a slightly wrong area, it produces a meaningless one.
Where the error comes from
- Your clicks. The result is a measurement of the points you placed, not of the curve. Nothing here can tell the difference between a point on the line and a point near it.
- The calibration. It is set from a small number of reference points, and any error in them scales every result. Check it by re-measuring a known feature of the plot before trusting the area.
- Point spacing. Trapezoidal error falls with the square of the spacing, so widely spaced points under-resolve curvature. A convex stretch is systematically over-estimated and a concave stretch under-estimated, and the bias does not cancel.
- The image. Source image quality, resolution, compression artefacts, axis interpretation, PDF rendering and browser scaling all move the pixels underneath your clicks.
- Extrapolation. The integral covers the range you clicked. It says nothing about the curve outside it, and a truncated tail can be most of the real area.
Area, calibration, point placement, image scaling, PDF rendering and exported results may be inaccurate, incomplete or misleading because of user input, source image quality, axis interpretation, browser rendering or implementation defects.
Never use this for
- Pharmacokinetic AUC, exposure or dosing decisions, or any clinical use.
- Regulatory submissions, compliance evidence, or anything a certification rests on.
- Legal, forensic or financial evidence, including reconstructing data from a published figure as though it were the data.
- Publishing digitised values without saying they were digitised, or without the original dataset behind them.
- Safety, production or engineering decisions where the area governs a design or an acceptance.
Verify independently
Validate important results against original data, calibrated instruments, trusted numerical software, published references, or a qualified professional before using them in engineering, scientific, clinical, legal, financial, safety or production decisions. Where the original data exists, use the original data; a figure is a picture of a dataset, not the dataset.
Calibration
Type linear-axis values, then optionally click the target icons to visually map them on the canvas. Scientific literals such as 1E5, 10^5, and 5x10^7 are accepted.
Data Series
Results (AUC)
A measurement of your clicks, not of the curve. Trapezoids between points placed by eye, scaled by your calibration, on linear axes only. Widely spaced points over-estimate a convex stretch and under-estimate a concave one, and the integral covers only the range you clicked. Validate anything that matters against the original data.