I have attached a new file with more arenas, some of them have a guarantee and most of them do not. Hope this makes yet again really appreciate your explanations, and it makes things a lot more clear. But if I say the two numbers should be different, the options decreases. you have many different answers all of them are correct. if I ask you for two number making the number 4. if you have limited constraints, there will be multiple points satisfying the optimal solution. Your final point why you have many solutions. It is ok, to keep non linear functions since you have limited number of variables.

Why your formula is not linear, because of the max function I guess. Multistart lets you start in different locations so you will be seeing more mountains and evaluate their heights (most probably one of them will be the max) population 100 is just a number, you can play with it, the larger the number of variables you need a bigger number obviously.

if you start from one area you will only see few mountains. Now, if your function is not convex, you need to visit every mountain you see. Convex (actually Concave in the case of maximization) that you know that your planet has only one summit and if you find it, it is the highest point (does this make sense?) Now if you visit some other plant, and you find a very high summit, can you be sure it is the highest summit? The answer is no. Kind is convex when it has one summit, and we know there is only one.įor example we know that the highest summit on planet earth is mount Everest. Lastly, could someone please explain to me how I find the optimal solution for this example if it is even possible? I also tried to change the changing variables to column L, instead of changing the percentages it should change the Revenue amount (I thought it would provide the same answer but it didn't), with the constraint that the SUM of 'Revenue' should be = 100 000 US. I tried to run Solver (GRG Nonlinear) with 'Use Automatic Scaling' ticked and then I ran it again having it unticked - I got different results, can someone explain whether it should be ticked or not? Third, the objective is to maximize the gross profit in cell Q16. Second, I have tried to set the changing variables to the percentage figures in column K, the constraint is of course that the sum of these percentages should equal to 100%.

If we choose to spend 0 US on an arena, we still have to pay the minimum guarantee.įirst of all, is this a Simplex LP problem or is it GRG Nonlinear or Evolutionary? All of these arenas have a different percentage regarding how much money they will get from showing the add (COGS%), and all of these arenas also have something called 'minimum guarantee', meaning that we have to pay them this amount if (Revenue * COGS%) < Minimum Guarantee (please see the formulas in order to understand the problem better). I can distribute that money however I want on ten different arenas where the add will show. Let's say that I am in advertising and that I get an order of 100 000 US. I just joined this community and I am hoping to get some knowledge regarding Excel Solver - I have a specific problem that I am hoping someone can provide some useful insights.

Excel solver function multiple times