Math Review #2 “I just got lost in thought. It was unfamiliar territory” “What happens if you get scared half to death twice?” Math Review Friday June 4 2003 A) Introduction a. Symbols b. Operations c. Central Tendencies B) Linear Algebra C) Correlation/Regression Analysis D) System of Equations: Linear/Quadratic E) Applied Calculus b. Operations Basic Math Review Why logarithms? Power and product rules: logb(xy) = logb(x) + logb(y) logb(xn) = nlogb(x) These rules motivated the introduction of logarithms (by Napier, in early 17th Century) and motivated their use in scientific computation until… computers! b. Operations Basic Math Review Why logarithms? 12 2 Example: Calculate 5 7 First use logs, then use log tables: y = 75 / 212 Log y = Log (218 / 75) http://www.sosmath.com/tables/logtable/logtable.html b. Operations Basic Math Review a) Solve for x: ln(ea) = bx b) Solve for y using common logarithms (base 10): y = 175 c) Find the exponent of 10 that solves for x: x2 = 5.5.10-12 Basic Math Review c. Central Tendencies The most commonly used descriptive statistics are measures of central tendency The sample mean (: pronounced “x bar”) is: x x n i 1 xi n Where Sxi represents the sum of all values in the sample and n represents the sample size Basic Math Review c. Central Tendencies Mean: arithmetic average Median: middle value of a set of values Mode: the data value that occurs most often Basic Math Review c. Central Tendencies Let’s assume we have a student population (n = 47) Frequency Distribution 14 12 Frequency 10 8 6 4 2 0 22 23 24 25 26 27 28 29 30 31 32 33 34 Age But what happens if we have an outlier (skewed distribution )? Basic Math Review c. Central Tendencies Let’s assume we have a real student population MPA ('04) - Age Freq. Distribution 10 Frequency 9 8 Frequency 7 6 5 4 3 2 1 0 20 22 24 26 28 30 32 Age 34 36 38 40 42 Basic Math Review B) Linear relationships Let’s play… Address# 1372 1585 1656 2252 2379 2681 1460 1841 2045 2600 2851 3410 2521 2929 Street # 37 48 51 81 87 103 42 61 71 99 111 139 95 115 Starbucks anyone? Basic Math Review B) Linear Relationships 160 140 Street Number 120 100 80 60 40 20 0 0 1000 2000 Broadway Address 3000 4000 Basic Math Review B) Linear Algebra The “slope” (m) of a line is its rate of change: 160 140 Slope: or (y2-y1)/(x2-x1) Street Number Dy/Dx 120 100 80 60 40 20 0 0 1000 2000 Broadway Address 3000 4000 Basic Math Review B) Linear Realtionships The “intercept” (b) of a line is the point where x = 0 160 140 Street Number 120 100 80 60 40 20 0 0 1000 2000 Broadway Address 3000 4000 Basic Math Review B) Linear Relationships The function f(x) = y = mx + b You can use it to make predictions 160 140 y = 0.05x - 31.317 Street Number 120 2 R = 0.9999 100 80 60 40 20 0 0 1000 2000 Broadway Address 3000 4000 Basic Math Review B) Graphing Linear Relationships Let’s assume we have a real fish population Weight (lb) 1.18 1.35 1.71 1.72 1.99 2.02 2.58 4.26 4.5 7.31 7.99 8.1 Length (in) 13.53 14.5 13.5 16.03 16.42 15.83 15.72 21.1 21.47 22.96 24.39 23.17 Any question regarding this data set? Basic Math Review B) Correlation The sample mean is: x x i n 1 xi n Sum of squares for variable x. This statistics quantifies the spread of variable x: n SSXX (x i x) i1 2 Basic Math Review B) Correlation Sum of squares for variable y. This statistics quantifies the spread of variable y: n SSYY (y i y) i1 2 Basic Math Review B) Correlation Sum of the cross-products. This statistics is analogous to the other sums of squares except that it quantifies the extent to which the two variables go together or apart: n SSXY (x i x)(x i x) i1 Basic Math Review B) Correlation Fish Data: Weight (lb) 1.18 1.35 1.71 1.72 1.99 2.02 2.58 4.26 4.5 7.31 7.99 8.1 Length (in) 13.53 14.5 13.5 16.03 16.42 15.83 15.72 21.1 21.47 22.96 24.39 23.17 SSxx: 78.5 SSyy: 182.0 SSxy: 113.8 The correlation coefficient is: SS XY r ( SS XX )( SSYY ) Here r = 0.95 Basic Math Review B) Correlation: Fish Data 25 Length (in) 13.53 14.5 13.5 16.03 16.42 15.83 15.72 21.1 21.47 22.96 24.39 23.17 20 Length (in) Weight (lb) 1.18 1.35 1.71 1.72 1.99 2.02 2.58 4.26 4.5 7.31 7.99 8.1 15 10 5 0 0 2 4 6 8 Weight (lb) The correlation coefficient is positive 10 Basic Math Review B) Correlation: the correlation coefficient has no inherent value, and in the exception of strong relationships as in the case presented, r is hard to use to determine correlational strength. Another statistics is much more useful: the coefficient of determination (r2) Weight (lb) 1.18 1.35 1.71 1.72 1.99 2.02 2.58 4.26 4.5 7.31 7.99 8.1 Length (in) 13.53 14.5 13.5 16.03 16.42 15.83 15.72 21.1 21.47 22.96 24.39 23.17 Basic Math Review B) Correlation: 25 Here r2 = 0.91 Length (in) 20 y = 1.4492x + 12.819 2 R = 0.9058 15 10 5 0 0 2 4 6 8 10 Weight (lb) This statistic quantifies the proportion of the variance of one variable that is explained by the other – Functional? Basic Math Review B) Linear Algebra Forgot a section of the fish data set Weight (lb) 0.015 0.05 0.06 0.07 0.08 0.09 0.12 0.15 0.16 0.25 0.27 0.33 0.42 0.44 0.5 0.53 0.6 0.83 Length (in) 3.16 6.07 5.72 6.57 4.32 5.52 8.39 8.32 7.79 6.05 8.11 8 10.13 10.97 9.72 11.02 11.33 13 Basic Math Review B) Correlation: 2 Here r = 0.82 15 Length (in) 12 9 y = 10.339x + 5.1588 2 R = 0.815 6 3 0 0 0.2 0.4 0.6 Weight (lb) 0.8 1 Basic Math Review B) Whole data set 25 20 y = 1.4492x + 12.819 Length (in) R2 = 0.9058 15 10 y = 10.339x + 5.1588 5 2 R = 0.815 0 0 2 4 6 Weight (lb) 8 10 Basic Math Review B) Correlation: Linear? 25 Length (in) 20 15 y = 2.2819x + 8.3152 10 R2 = 0.8172 5 0 0 2 4 6 Weight (lb) 8 10 Basic Math Review B) Non-linear relationship 25 Length (in) 20 15 10 5 0 0 2 4 6 Weight (lb) 8 10 Basic Math Review B) Non-linear relationship Let’s make a statements about the relationship: -) The weight is to the volume WV Where: V=AxL A = a x L2 V = a x L3 W=rxV Therefore W = a x r x L3 25 Length (in) 20 15 10 5 0 0 2 4 6 Weight (lb) 8 10 Basic Math Review B) Non-linear relationship W = a x r x L3 L W 3 L W 1 ar L k 3 W kW ar 25 20 Length (in) 3 1 1 3 15 10 5 0 0 2 4 6 Weight (lb) 8 10 Basic Math Review B) Non-linear relationship 1 L3 W ar 1 L k 3 W kW ar 25 20 Length (in) L W 3 15 10 0.307 y = 12.797x 2 R = 0.9462 5 0 0 2 4 6 Weight (lb) 8 10 1 3 L k 3 W kW Log L = Log (k x W1/3) Log L = Log k + 1/3 Log W y = b + mx 1 3 1.6 Length (in) 1.2 0.8 y = 0.307x + 1.1071 2 R = 0.9462 0.4 0.0 -2.0 -1.6 -1.2 -0.8 -0.4 Weight (lb) 0.0 0.4 0.8 Monday Lamont orientation (LDEO Exec. Director and DEES Chair) Math Review #3: System of Equations: Linear/Quadratic - Applied Calculus