The transformation relies on the Product Rule of logarithms: . This allows the separation of a constant coefficient from the variable term.
The Power Rule of logarithms, , is essential for moving the variable from the exponent (in exponential models) or the power (in power models) into a linear multiplier position.
These laws apply regardless of the base used (common log or natural log ), provided the base is consistent throughout the calculation.
For an exponential model of the form , taking the natural logarithm of both sides yields .
This matches the linear equation where the vertical axis represents and the horizontal axis represents .
The gradient () of the resulting line is , and the vertical intercept () is . To find the original constant , one must calculate .
For a power model of the form , taking the logarithm of both sides yields .
In this case, the linear form is achieved by plotting on the vertical axis against on the horizontal axis.
The gradient () of the line directly represents the power , while the vertical intercept () represents . The constant is recovered using the inverse log: .
| Feature | Exponential Model () | Power Model () |
|---|---|---|
| Vertical Axis | ||
| Horizontal Axis | ||
| Gradient () | ||
| Intercept () |
The critical difference lies in the horizontal axis: exponential models keep the independent variable in its original form, whereas power models require the logarithm of .
If a plot of against is a straight line, the relationship is exponential. If a plot of against is a straight line, the relationship is a power law.
Identify the Axes First: Before performing any calculations, check the labels on the axes. If only the y-axis is logarithmic, it is an exponential model. If both axes are logarithmic, it is a power model.
Base Consistency: Always use the same base for the gradient and the intercept. If the graph uses , use to find the constants; if it uses , use .
Intercept Interpretation: The intercept is the value of the log-variable, not the constant itself. You must always perform the inverse operation (e.g., ) to find the physical constant.
Units and Scale: Ensure that any units provided in the problem are handled before taking logarithms, as you cannot take the log of a value with units (logs are dimensionless).
Mixing up and : Students often plot against for a power model, which results in a curve rather than the intended straight line.
Negative Values: Logarithms are only defined for positive numbers. If the data contains zero or negative values, a simple log transformation cannot be applied without first shifting the data.
Gradient Miscalculation: For exponential models , the gradient is exactly . For , the gradient is . Forgetting which form is being used leads to incorrect constant values.