Correlation and Regression Calculator

Instructions

1. Paste a table with a header and at least two numeric pairs.
2. Choose X and Y columns and calculate means, Pearson r, slope, intercept and R squared.
3. The fitted line is Y = slope × X + intercept.
4. Download the statistics as CSV.

Understanding your result

The sample follows Y = 2X + 1 exactly, so the fitted slope is 2, intercept is 1 and correlation is positive and perfect. Real data can depart from a line even with a high correlation. Review the source pairs before interpreting the statistics.

What to do if you have a problem

If calculation is refused, check missing values and whether either column is constant. A single misplaced decimal or outlier can change the fit. Test with correctly paired observations; sorting one column independently would break the relationship you are trying to explore.

What to know before using the result

All pairs must contain finite numeric values with dot decimals; no missing pairs are silently dropped. Constant X or Y is rejected because correlation is undefined. Uses JavaScript floating-point arithmetic and a linear model with an intercept. Correlation does not establish causation. Outliers, nonlinear patterns and small samples can mislead; no confidence intervals, hypothesis tests or predictive guarantees are provided. Input limit: 2 MB and 20,000 rows including the header.

These tools work with pasted text or supported CSV files rather than Excel workbook files. CSV exports contain values and do not preserve workbook formatting, formulas or multiple worksheets. Formula builders produce formula text to use in Excel. Data processing runs in the browser tool.

Try an example

X,Y
1,3
2,5
3,7
4,9

Summary

Review the fitted relationship alongside the source pairs, outliers and limits of a linear model.