Summary of "AP Precalculus Unit 2 Summary Review - Exponential and Logarithmic Functions"

Overview

This summary covers the main ideas, methods, and worked examples from the video. It has been edited for clarity and to correct obvious subtitle errors.

Big-picture topics covered

Sequences (arithmetic and geometric) and their relation to linear and exponential functions
Exponential functions: form, graphs, behavior, concavity, and exponent rules
Logarithmic functions: definition as inverse of exponentials, rules, solving log equations, domain/extraneous-solution checks
Solving exponential and logarithmic equations (methods)
Function composition
Inverse functions (table/graph/analytic), domain restrictions, and examples including rational-function inverses
Regression models: fitting (especially exponential), residuals, residual-plot diagnostics, secant lines, average rate of change, and using a secant line to make/predict values

Key formulas and equivalences

Arithmetic sequence (general): a_n = a_k + d(n − k) (equivalently a_n = a_1 + (n − 1)d)
- d = common difference (subtract consecutive terms)
- domain: n ∈ {1, 2, 3, …} (positive integers)
Geometric sequence: g_n = g_k · r^(n − k) (equivalently a·r^(n−1))
- r = common ratio (divide consecutive terms)
- domain: n ∈ {1, 2, 3, …}
Exponential function: y = A·B^x
- B > 1 → growth; 0 < B < 1 → decay
- A < 0 → reflection across x-axis
- Horizontal asymptote for basic a·b^x: y = 0
Logarithm-exponential equivalence: log_b(W) = X ⇔ b^X = W
Residual: residual = actual y − predicted y

Methodologies — step-by-step procedures

Build an arithmetic-sequence formula from a known term
- Identify the common difference d (subtract consecutive terms).
- Choose a known term a_k at position k.
- Use a_n = a_k + d(n − k). Plug numeric a_k, d, and k to get a formula for a_n.
- Domain: n must be positive integers (check by plugging n = 1,2,…).
Build a geometric-sequence formula from a known term
- Identify the common ratio r (divide consecutive terms).
- Choose a known term g_k at index k.
- Use g_n = g_k · r^(n − k). Optionally rewrite as A·B^n by algebraic manipulation.
- Domain: n must be positive integers.
Recognize exponential functions from input–output behavior
- Inputs change additively (equal-length intervals) while outputs change multiplicatively (multiply by a constant ratio).
- In tabular data, look for a constant multiplicative change in successive outputs.
Analyze exponential-graph behavior and concavity
- If B > 1 and A > 0: increasing growth, concave up; limits: x→∞ is ∞, x→−∞ is 0.
- If B > 1 and A < 0: reflected (negative) growth in magnitude, concave down.
- If 0 < B < 1 and A > 0: decay toward 0 as x→∞; concavity may be up if decreases slow.
- If 0 < B < 1 and A < 0: reflected decay, concave down.
- For basic a·b^x: horizontal asymptote y = 0 and no x-intercepts.
Common exponent rules (for simplification)
- b^m · b^n = b^(m+n)
- b^m / b^n = b^(m−n)
- (b^m)^n = b^(m·n)
- b^(−n) = 1 / b^n
- Strategy: rewrite factors with a common base, then combine exponents.
Convert an exponential expression to standard A·B^x form
- Use exponent rules to separate factors, move constant factors to numerator/denominator to isolate B^x, simplify numeric constant to get A.
Solving exponential equations (two approaches)
- If both sides can be written with the same base, set exponents equal (e.g., 2^(x+2) = 2^5 → x+2 = 5).
- Otherwise isolate the exponential and take logarithms: B^(expression) = C → expression = log_B(C) (or use natural log and change-of-base).
Logarithm basics and rules
- log_b(xy) = log_b(x) + log_b(y)
- log_b(x/y) = log_b(x) − log_b(y)
- log_b(x^k) = k·log_b(x)
- log_b(W) = X ⇔ b^X = W
Solving logarithmic equations
- Combine logs using log rules to get a single log when possible.
- If log_b(A) = log_b(B), then A = B (provided arguments are positive).
- Convert single-log equations to exponential form when needed.
- Always check domain: log arguments must be positive; discard solutions that make arguments ≤ 0.
Function composition (f ∘ g) - (f ∘ g)(x) = f(g(x)); evaluate g(x) first, then apply f. - Two approaches: form f(g(x)) algebraically then evaluate, or evaluate g at x and then f at that result.
Finding inverse functions - From a table: swap x and y columns; ensure each new input gives exactly one output (otherwise not invertible without domain restriction). - Analytically: set f(x) = y, swap x and y, solve for y, then rename y as f^−1(x). - If solving yields ± (two values), the inverse is not a function unless you restrict the domain (choose one branch). - For rational functions: after swapping, clear denominators, collect y terms, factor y, and divide to isolate y; then note domain restrictions.
Regression models and residuals - Goal: choose a model (linear, exponential, quadratic, log) to fit scatter data for prediction. - Exponential model form: y = A·B^x. - Residual = actual y − predicted y: - Positive residual: model underestimates (point above model). - Negative residual: model overestimates (point below model). - Residual-plot diagnostic: - Randomly scattered residuals (no pattern) → model appropriate. - Systematic patterns (e.g., oscillating residuals) → chosen model is inappropriate. - Building an exponential model from two points (x1, y1) and (x2, y2): - y1 = A·B^(x1), y2 = A·B^(x2). - Divide: y2 / y1 = B^(x2 − x1). - Solve for B: B = (y2 / y1)^(1/(x2 − x1)). - Then A = y1 / B^(x1). Round sensibly when using a calculator. - Average rate of change between x1 and x2: (y2 − y1) / (x2 − x1) (slope of secant line). - Use the secant-line equation (point-slope) for linear predictions; note that secant-line predictions may differ from the exponential model and can over/underestimate differently inside vs. outside the interval.

Worked-example highlights

Arithmetic example: given 4th term = 7 and d = 5, derived a_n = 5n − 33 and checked by plugging in n values.
Geometric example: terms multiply by 3 and 3rd term = 36 → g_n = 36·3^(n−3); converted to A·B^n form to get (4/3)·3^n.
Exponent-rule simplification: combined multiple bases to a single base 2 and obtained final exponent 5 − 10x after algebra.
Solving an exponential equation: isolated and rewrote 32 = 2^5 to get x = 3 (or used log base 2).
Solving a logarithmic equation: combined logs to a single log, solved the resulting quadratic, and checked domains; both roots were valid.
Inverse of a rational function: swapped x and y, cleared denominators, collected and factored y, then isolated y to get f^−1(x) = (2 + 4x)/(3x − 1) (with domain considerations).
Regression from two points: with (3,10) and (7,42) computed B = (42/10)^(1/4) ≈ 1.1432, A ≈ 3.45; average rate of change = 8; used secant line for predictions and discussed over/underestimates relative to the exponential model.

Important exam-focused takeaways (AP Precalculus)

Know the difference between arithmetic vs geometric sequences and how they correspond to linear vs exponential models. Remember sequence domains are positive integers.
Be fluent with exponent and logarithm rules — they commonly appear on FRQs.
Know both methods for solving exponential equations: rewriting with a common base and using logarithms.
Always check the domain when solving logarithmic equations; discard any solutions that make log arguments nonpositive.
Composition and inverse functions are commonly tested (including with tables and graphs).
For regression tasks: know how to derive an exponential model from two points, compute residuals, interpret residual plots (no pattern = appropriate model), compute average rate of change and secant-line equations, and reason about over/underestimates.