Exponential Regression Calculator

The exponential regression model is:

$$y = a \cdot e^{bx}$$

Taking the natural logarithm of both sides linearizes the equation:

$$\ln(y) = \ln(a) + bx$$

This is a linear equation in the form \(Y' = A + Bx\), where \(Y' = \ln(y)\), \(A = \ln(a)\), and \(B = b\). Standard linear regression is then applied to the transformed data points \((x_i, \ln(y_i))\) to find the slope B and intercept A using the ordinary least squares formulas:

$$b = B = \frac{n\sum x_i \ln(y_i) - \sum x_i \sum \ln(y_i)}{n\sum x_i^2 - (\sum x_i)^2}$$

$$\ln(a) = A = \overline{\ln(y)} - b\bar{x} \quad \Rightarrow \quad a = e^A$$

The parameter b determines the nature of the curve: when b > 0, the function represents exponential growth (Y increases as X increases); when b < 0, it represents exponential decay (Y decreases toward zero). The parameter a is the Y-value when X = 0 (the initial condition). The doubling time for growth is \(t_d = \ln(2)/b\), and the half-life for decay is \(t_{1/2} = \ln(2)/|b|\).

The R² value is calculated on the original (untransformed) scale by comparing the sum of squared residuals from the exponential predictions to the total sum of squares, providing a direct measure of how well the exponential curve fits the actual data.

Understanding your exponential regression results:

• Coefficient a: The Y-intercept of the exponential curve (value at X = 0). For population models, this is the initial population. For decay, it is the initial quantity.
• Growth rate b: The continuous growth/decay rate. A b of 0.5 means a 65% increase per unit X (since e⁰·⁵ ≈ 1.65). A b of -0.3 means a 26% decrease per unit X.
• R²: Goodness of fit on the original scale. High R² confirms the data follows an exponential pattern. If R² is low, the relationship may be linear, polynomial, or follow a different nonlinear form.

Important considerations: (1) All Y values must be positive for the logarithmic transformation. (2) The linearization gives equal weight to all transformed residuals, which may not minimize residuals on the original scale optimally — for precise work, nonlinear least squares is preferred. (3) Exponential models grow without bound, so extrapolation should be done cautiously. Real-world growth eventually encounters limiting factors (logistics, resources, saturation).