Free online tool
Make a linear calibration curve
Fit a linear standard curve from known concentrations and measured responses. Review residuals, keep replicates, and export figures and a calculation report.
Enter or ↑ ↓ moves between rows. Alt + ← → moves between columns. Tab moves through fields.
| # | ||
|---|---|---|
| 1 | ||
| 2 |
Use one row per measurement. Concentration must be nonnegative; response may be negative. Include column headings.
CSV, TSV, JSON rows, JSON Lines or values-only XLSX, up to 5 MB. Choose a worksheet before import. All processing stays in this browser.
Import settings
Column mapping is available after opening the data. JSON object keys become column names. Grid column names are separate from its data rows.
- Import or paste standards with known concentrations and measured responses.
- Map concentration and response columns, add units and review the fitted line and residuals.
- Export figures and a calculation report, or save an editable project.
- Slope
- 1.008
- Intercept
- 0.093333
- R²
- 0.99937
6 fictional standards · review residuals before interpreting the fit
From input to finished work
How it works
Turn known standards into an ordinary least-squares line with an intercept. Keep every replicate, review fitted values and residuals, and save the original measurements with your project. Processing stays in this browser.
No account needed. Your input stays in this browser. Export your finished work to keep a portable copy.
A little guidance
Useful things to know
Input tips, checks and limits for make a linear calibration curve.
What kind of calibration curve can I make?
The app fits an unweighted linear ordinary least-squares model with an intercept. It requires at least three included measurements and two distinct concentrations. Replicates stay as separate rows.
Can I use a blank standard?
Yes. Enter a concentration of zero for a measured blank. An empty cell is missing data and must be corrected or explicitly excluded with a reason.
Does the app calculate unknown concentrations or detection limits?
No. This version fits the forward response model and reports residuals. Inverse prediction, detection limits, uncertainty intervals and weighted or nonlinear fits require separately validated methods.
What do residuals tell me?
Residuals are observed response minus fitted response. Their pattern can reveal departures from a straight-line model; a high R² alone does not establish that calibration is appropriate.
Are my measurements uploaded?
Measurements are processed locally in your browser. You can save a project locally, clear it or export figures and a calculation report.