Calibration curves

Each quant has a curve. View → Samples and results → Calibration shows the fit for the sample you are on.

Curve types

The Fit column is one of:

Fit	What it does
Linear	Straight line, y = a + b·x.
Quadratic	y = a + b·x + c·x².
Average of response factors	Mean of (response / concentration). Used for every ISTD quant.
Power	y = a·x^b.
First order ln	First-order fit in natural-log space.
Second order ln	Second-order fit in natural-log space.
Point to point	Straight segments between adjacent calibration doses. Points outside the calibrated range use only the end dose and are forced through the origin.

Origin

Origin	What it does
Force	The line goes through (0, 0). Zero intercept.
Ignore	The intercept is fitted. No (0, 0) point is added.
Include	(0, 0) is included as a calibration point.
Blank offset	The intercept comes from the blanks selected for that quant.

Weighting

Weight	What it does
Equal	Every point has the same weight.
1/x	Weight is 1/x.
1/x^2	Weight is 1/x².
1/y	Weight is 1/y.
1/y^2	Weight is 1/y².
Log	Logarithmic weight.
1/SD^2	Weight is 1/σ² (inverse variance).

x is concentration. y is response.

Defaults

A new quant that is not an internal standard is Linear, origin Force, weighting Equal. An ISTD quant is stored as Average of response factors. Opening an existing method does not rewrite the fit, origin, or weight already saved on a quant.

Fit quality

Each bracket stores:

R² (shown as R² on the curve). A limit group can set a minimum R². If the bracket’s R² is missing or below that minimum, the result’s R² flag is Too low. The fit itself is shared; each sample’s limit group supplies its own cutoff. See Limit groups.
Relative residual on each calibration point, and the maximum relative residual (and the dose where it occurred) on the bracket.

Calibration standards with a recovery outside the limit-group window are also flagged. Those flags are separate from R². See Checks.

Example curves for each fit, origin, and weight are in Example batches.