📚 Logarithms and Non-linear Data | 对数与非线性数据
In Edexcel A Level Mathematics, logarithms are not only tools for solving exponential equations; they also allow us to transform non-linear relationships into straight lines. This topic appears in both pure and applied contexts, especially when dealing with experimental data that follows an exponential or power law. Understanding how to linearise data and interpret transformed regression lines is essential for success in the statistics and mechanics components.
在 Edexcel A Level 数学中，对数不仅是求解指数方程的工具，还能把非线性关系转化为直线。这个主题既出现在纯数学部分，也出现在应用部分，尤其是处理服从指数律或幂律的实验数据时。理解如何将数据线性化并解释变换后的回归线，对统计和力学模块的成功至关重要。
1. Why logs matter in data analysis | 为什么数据分析需要对数
Many real-world quantities do not grow at a constant rate; populations, radioactive decay, prices and learning curves often follow exponential or power relationships. Plotting raw data from these situations usually gives a curve, which makes it hard to estimate parameters or compare models. Taking logarithms can convert multiplication and powers into addition and multiplication, so a curved relationship becomes linear.
许多现实世界的量并不是以恒定速率变化的；人口、放射性衰变、价格和学习曲线通常遵循指数或幂关系。绘制这些情境下的原始数据通常会得到曲线，这使得估计参数或比较模型变得困难。取对数可以把乘法和幂转化为加法和乘法，因此曲线关系就变成了线性关系。
Once data are linearised, we can use the familiar least squares regression line to model the transformed data. This gives estimates of the constants in the original non-linear model, and also allows us to assess the goodness of fit using standard linear regression ideas.
一旦数据被线性化，我们就可以使用熟悉的最小二乘回归线来拟合变换后的数据。这样可以得到原始非线性模型中常数的估计值，并允许我们使用标准线性回归的思想来评估拟合优度。
2. Logarithm laws recap | 对数运算律回顾
rmula is logₐx = log x / log a, where the logs on the right can be in any consistent base.
在变换数据之前，你必须熟练掌握对数运算律。对于任意正底数 a，logₐ(xy) = logₐx + logₐy，logₐ(x/y) = logₐx − logₐy，logₐ(xⁿ) = n logₐx。换底公式为 logₐx = log x / log a，其中右边的对数可以使用任意一致的底数。
| Law | 法则 | Expression | 表达式 |
|---|---|
| Product | 乘法 | logₐ(xy) = logₐx + logₐy |
| Quotient | 除法 | logₐ(x/y) = logₐx − logₐy |
| Power | 幂 | logₐ(xⁿ) = n logₐx |
| Change of base | 换底 | logₐx = log x / log a |
3. Exponential models: y = abˣ | 指数模型 y = abˣ
An exponential model has the form y = abˣ tity grows; if 0 < b < 1 it decays. Taking logs to any base gives log y = log a + x log b.
指数模型的形式为 y = abˣ，其中 a 是初始值，b 是增长或衰减因子。如果 b > 1，量增长；如果 0 < b < 1，量衰减。对两边取任意底数的对数，得到 log y = log a + x log b。
erefore a graph of log y against x should be a straight line with gradient log b and intercept log a.
该方程具有 Y = mX + c 的形式，其中 Y = log y，X = x，m = log b，c = log a。因此，绘制 log y 对 x 的图形应该是一条直线，斜率为 log b，截距为 log a。
4. Power models: y = axⁿ | 幂函数模型 y = axⁿ
a + n log x. Now the independent variable is log x, not x, so we plot log y against log x.
幂函数模型的形式为 y = axⁿ，其中 a 是常数，n 是幂。取对数得到 log y = log a + n log x。此时自变量是 log x 而不是 x，因此我们绘制 log y 对 log x 的图形。
The gradient of the resulting straight line is n and the vertical intercept is log a. This is the standard method for recognising power laws in experimental data, for example in physics when relating period and length of a pendulum.
所得直线的斜率是 n，纵截距是 log a。这是识别实验数据中幂律的标准方法，例如在物理学中处理单摆周期与摆长的关系时。
5. Linearising by taking logs | 取对数线性化
use X = log x and Y = log y. Once transformed, all the standard regression formulas apply.
线性化是指变换变量，使关系变为 Y = mX + c。对于指数型数据，使用 X = x，Y = log y；对于幂型数据，使用 X = log x，Y = log y。变换后，所有标准回归公式都适用。
It is important to plot the transformed data first. If the points roughly lie on a straight line, the model is appropriate. If the transformed data still curve, the chosen model is likely wrong and a different transformation or model should be tried.
首先绘制变换后的数据非常重要。如果点大致落在一条直线上，说明模型合适。如果变换后的数据仍然是弯曲的，那么所选模型很可能不对，应该尝试不同的变换或模型。
6. Using log graphs and semi-log plots | 对数图与半对数图
A semi-log plot has one logarithmic axis and one linear axis. Plotting log y against x is equivalent to using a logarithmic vertical axis and a linear horizontal axis. This is useful for exponential data because an exponential curve becomes a straight line.
半对数图有一个对数轴和一个线性轴。绘制 log y 对 x 的图形相当于使用对数纵轴和线性横轴。这对指数数据很有用，因为指数曲线会变成一条直线。
A log-log plot has both axes logarithmic, so plotting log y against log x is equivalent to using logarithmic scales on both axes. Power laws become straight lines on a log-log plot, with the gradient giving the power n.
双对数图的两个轴都是对数刻度，因此绘制 log y 对 log x 的图形相当于在两个轴上都使用对数刻度。幂律在双对数图上变成直线，斜率给出幂 n。
7. Least squares regression with transformed data | 变换数据的最小二乘回归
ion, calculate the transformed values for each data point. For an exponential model y = abˣ , form the pairs (log x, log y). Then compute the least squares regression line of Y on X.
选择变换后，为每个数据点计算变换值。对于指数模型 y = abˣ，构造数据对 (x, log y)；对于幂模型 y = axⁿ，构造数据对 (log x, log y)。然后计算 Y 对 X 的最小二乘回归线。
= Σ(X − mean X)². These statistics are exactly the same as in ordinary linear regression.
回归线方程为 Y = mX + c，其中 m = SXY / SXX，c = mean(Y) − m mean(X)。这里 SXY = Σ(X − mean X)(Y − mean Y)，SXX = Σ(X − mean X)²。这些统计量与普通线性回归完全相同。
8. Estimating parameters from the regression line | 从回归线估计参数
Once the regression
Published by TutorHao | A-Level Mathematics Revision Series | aleveler.com
更多咨询请联系16621398022（同微信）
或截图后用「扫一扫 → 相册」选择本图