1 / 27

Exponentials, logarithms and rescaling of data

Exponentials, logarithms and rescaling of data. Math 151 Based upon previous notes by Scott Duke-Sylvester as an adjunct to the Lecture notes posted on the course web page. Definitions. f(x) = a x , exponential function with base a f(x) = log a x, logarithm of x base a

pearl
Télécharger la présentation

Exponentials, logarithms and rescaling of data

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. Exponentials, logarithms and rescaling of data Math 151 Based upon previous notes by Scott Duke-Sylvester as an adjunct to the Lecture notes posted on the course web page

  2. Definitions • f(x) = ax, exponential function with base a • f(x) = logax, logarithm of x base a • f(x) = axb, is an allometric function

  3. Motivation • Many biological phenomena are non-linear: • Population growth • Relationship between different parts/aspects of an organism (allometric relationships) • The number of species found in a given area (species-area relationships) • Radioactive decay and carbon dating • Many others • Exponentials, logs and allometric functions are useful in understanding these phenomena

  4. Population growth is a classic example • Algae : cell division • Geometric growth 32 64 … 8 16 4 2 1 t = 0 1 2 3 4 5 6

  5. Exponentialsf(x) = ax, a > 0 0 < a < 1 exponential decrease a > 1 exponential increase As x becomes very negative, f(x) gets close to zero As x becomes very positive, f(x) gets close to zero

  6. Special case, a = 1

  7. f(x) = ax is one-to-one. For every x value there is a unique value of f(x). • This implies that f(x) = ax has an inverse. • f-1(x) = logax, logarithm base a of x.

  8. f(x) = ax f(x) = logax

  9. logax is the power to which a must be raised to get x. • y = logax is equivalent to ay = x • f(f-1(x)) = alogax = x, for x > 0 • f-1(f(x)) = logaax = x, for all x. • There are two common forms of the log fn. • a = 10, log10x, commonly written a simply log x • a = e = 2.71828…, logex = ln x, natural log. • logax does not exist for x ≤ 0.

  10. Laws of logarithms • loga(xy) = logax + logay • loga(x/y) = logax - logay • logaxk = k·logax • logaa = 1 • loga1 = 0 • Example 15.7 :

  11. A radioactive material decays according to the law N(t)=5e-0.4t When does N = 1? For what value of t does N = 1? • Example 15.8 : Radioactive decay

  12. To compute logax if your calculator doesn’t have loga or use Example:

  13. general formula for a simple exponential function: • f(x) = x or • y = x • Then • ln(y) = ln x • ln(y) = ln  + ln (x ) • ln(y) = ln  + x ln  • Let b = ln , and m = ln , Y=ln (y) then this shows • Y = b + mx • which is the equation of a straight line. • This is an example of transforming (some) non-linear data so that the transformed data has a linear relationship. An exponential function gives a straight line when you plot the log of y against x (semi-log plot). Special exponential form : f(x) = emx

  14. Consider the algae growth example again. How do you know when a relationship is exponential? Regular plot Semilog plot N= t ln(N) = mt+b

  15. ln(N) = (0.693)t • Fit a line to the transformed data • Estimate the slope and intercept using the least squares method. • Y=mx+b • b~0, m = 0.693.. • Estimate  and . • b = ln ->  = eb = e0 = 1 • m = ln ->  = em = e0.693.. = 2.0 • N = 2t N=2t

  16. Example 17.9 : Wound healing rate Regular plot Semilog plot

  17. How do you make a semilog plot? • Use the semilog(x,y) command in Matlab • Take the log of one column of data and plot the transformed data (here log y) against the untransformed data (here x)

  18. ln(A)=4.96-0.048t • Estimate slope and intercept using least squares • Y = b+mx • m = -0.048 • b = 4.69 • b = ln ->  = eb = e4.69 = 108.85 • m = ln ->  = em = e-0.048 = 0.953 • A = 108.85(0.953)t A=108.85(0.953)t

  19. f(x)=bxa • Allometric relationships (also called power laws) • Describes many relationships between different aspects of a single organism: • Length and volume • Surface area and volume • Body weight and brain weight • Body weight and blood volume • Typically x > 0, since negative quantities don’t have biological meaning.

  20. a > 1 a < 0 a = 1 a < 1

  21. f(x) = bax f(x) = bxa

  22. Example : It has been determined that for any elephant, surface area of the body can be expressed as an allometric function of trunk length. • For African elephants, a=0.74, and a particular elephant has a surface area of 20 ft2 and a trunk length of 1 ft. • What is the surface area of an elephant with a trunk length of 3.3 ft? • x = trunk length • y= surface area • y = bxa = bx0.74 • 20 = b(1)0.74 20 = b • y=20x0.74 • y=20(3.3)0.74=48.4 ft2

  23. How do you know when your data has an allometric relationship? • Example 17.10

  24. Regular plot log-log plot Semilog plot

  25. How do you make a log-log plot? • Use the loglog(x,y) command in matlab • Take the log of both columns of data and plot the transformed columns.

  26. Y=bxa • ln(y)=ln(bxa) • ln(y) = ln(b) + ln(xa) • ln(y) = ln(b) + a ln(x) • Let • Y = ln(y) • X=ln(x) • B=ln(b) • Then • Y=B + a X • Which is the equation for a straight line. So an allometric function gives a straight line on a loglog plot.

More Related