Professional Documents
Culture Documents
1
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 A B C
Solution for questions (a, b, c): If you have not already done so, start @Risk. 1. Make Demand a normally distributed random variable as follows: Select Cell C11. Pull down the "Insert" menu and select "Function" In the left window, select "@Risk Distribution" (see picture below). In the right window, select "RiskNormal" Click OK Enter the mean and standard deviation of demand. Click OK Cell C11 should now read: =RiskNormal(500,20).
Abbreviation for the preceding instructions: Select C11. Menu: Insert, Function, Left Window: @Risk Distribution, Right Window: RiskNormal, OK, Mean: 500. std dev: 20. OK. 2. Make Profit an "output" so @Risk will give you a report on it: Select C13. Menu: @Risk, Model, Add Output Cell C13 should now read: =RiskOutput() + 7*C12 - 5*C10 3. Set @Risk to simulate Demand 1000 times: Menu: @Risk, Simulation, Settings, Iterations tab. # Iterations 1000, # Simulations 1. OK.
71 6. Three new worksheets will appear. Their contents are shown below. 72 Answer (a): Profit averages $888.36 with standard deviation $121.30 73 Answer (b): The distribution (graph) is bell shaped, except for a spike on the right side. 74 Answer (c): The Output Statistics Report shows that profits are below $800 about 24% of the time. 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 Input Statistics Report
Inputs Demand Simulation# 1 Statistics / Cell $C$11 Minimum 436.647 Maximum 578.549 Mean 500.012 Standard Deviation 20.0344 Variance 401.377 Skewness 0.02406 Kurtosis 3.04675 NumErrs 0 Mode 494.674 5% 467.072 10% 474.327 15% 479.262 20% 483.103 25% 486.47 30% 489.495 35% 492.249 40% 494.888 45% 497.448 50% 499.981 55% 502.511 60% 505.048 65% 507.68 70% 510.448 75% 513.463 80% 516.831 85% 520.686 90% 525.606 95% 532.771
Output Graphs
155 Answer (d): Average profit is highest in simulation 2, for which Q=480. 156 Based on these results, you can try a "finer grid" of values for Q near the best one. 157 158 Output Name Output Cell Simulation# Minimum Maximum Mean 159 Profit $C$13 1 765.781555 920 918.8639 160 2 665.781555 960 948.3878 161 3 565.781555 1000 944.1999 162 4 465.781525 1040 888.3872 163 5 365.781525 1080 798.8599 164 6 265.781525 1120 699.9896 165 7 165.78154 1053.043 600.0335 166 167 Input Name Input Cell Simulation# Minimum Maximum Mean 168 Stock (Q) $C$10 1 460 460 460 169 2 480 480 480 170 3 500 500 500 171 4 520 520 520 172 5 540 540 540 173 6 560 560 560 174 7 580 580 580
Std Dev
0 0 0 0 0 0 0
43 Simulations 1. 44 OK. 45 46 47 48 49 50
101 102 103 w: RiskSimtable. OK. 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150
2 This is the same example, illustrating how to "fit" some distributions. 3 Two @Risk functions use "tables" as input: 4 >> RiskGeneral randomly generates values anywhere between two limits, on a continuous scale. 5 >> RiskDiscrete randomly generates values from the top row of the table. 6 The other distributions are self-explanatory. (See the note below about the LogNormal distribution.) 7 8 Continuous Discrete Table Normal Log Expo Uniform Table Poisson Binomial Uniform 9 normal nential 10 Stock (Q) 520 520 520 520 520 520 520 520 520 11 Sales 500 500 500 500 500 500 500 500 500 12 Profit 900 900 900 900 900 900 900 900 900 13 Demand 500 500 500 500 500 500 500 500 500 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 X: Freq: Historical Frequencies of Demand: 460 480 500 520 540 22 Cells: Continuous Distributions C13 D13 E13 F13 G13 46 77 35 10
Formulas: = RiskGeneral(C15, G15, C15:G15, C16:G16) = RiskGeneral(Min, Max, Values of X, Frequencies) = RiskNormal(500,20) = RiskNormal(Mean, Standard Deviation) = RiskLognorm(500,20) = RiskLognorm(Mean, Standard Deviation) = RiskExpon(500) = RiskExpon(Mean) = RiskUniform(460,540) = RiskUniform(Lower Limit, Upper Limit)
Discrete Distributions
I13 = RiskDiscrete(C15:G15, C16:G16) = RiskDiscrete(Values of X, Frequencies) J13 = RiskPoisson(500) = RiskPoisson(Mean) K13 = RiskBinomial(1000,0.5) = RiskBinomial(n, p) L14 = RiskIntUniform(460,540) = RiskIntUniform(Lower Limit, Upper Limit)
Note about the Lognormal Distribution Lognormal has the shape of an asymmetrical bell. X cannot be negative, so the distribution begins at X=0. However, it has a long tail to the right. All positive X-values are possible. It is often used in place of the normal distribution to represent variables that cannot have negative values.