You need to perform operations research on a budget. What are the most affordable decision analysis tools?

It’s helpful but has limited domain to operate with. However, working on advanced spreadsheet tools need niche skill set. Working with basic models is easy and quick. Concepts like simulation can be operated through spreadsheet but more robust solutions can be performed better on other platforms like MATLAB

If you’re on a budget, spreadsheet software are your first choice as an operations research analyst to solve problems. Software such as Microsoft Excel allow you to visualize the data and have a rich set of functions where you can perform some modeling if you know a little Visual Basic. The solver in Excel can solve some small optimization problems. The strengths are data visualization and ease of use while the weaknesses are scalability and having a robust statistical modeling capability; if which case you may need a statistical package like R.

Spreadsheet software for Operation Research (on a budget) should be the top of mind of every Analyst, because it is the basic platform for you to at least doing analysis even if you haven't master the functionality of Spreadsheet you might able to come up with something using manual work / without formula and calculations that is how versatile Spreadsheet is, however when it comes to scalability then there is a problem since Spreadsheet should only be used for small scale research, one thing most people do is making Spreadsheet the data warehouse for their work, which is wrong. Overall Spreadsheet is a go to software to start your research or analysis (on a budget) without needing high end skills.

An operations research project is not just about mathematical modeling but also entails explaining/validating your methodology, communicating with stakeholders and demonstrating business value. Spreadheets are a very good choice initially to brainstorm and communicate with stakeholders since you can demonstrate how a constraint / objective contributes towards solving a business problem and get their buy-in. MS Excel with Opensolver add-in is a great combo. Later, to solve problem at larger scale you could move to Python + CBC solver combo (you could also look at Google OR Tools for VRPs, and IPOPT if problem is non-linear). I prefer Pyomo for modeling as you can easily switch to paid solver like Gurobi without any code change.

I agree with this. When the project has little to no budget on operations research, it is very wise to exploit the most accessible and available tools such as spreadsheets. In fact, there are many spreadsheet apps such as Microsoft Excel and Google Sheet. There you can perform a lot of calculations to support your operations research project. You may use existing feature such as formulas and built-in optimization tools such as Excel Solver.

I wouldn't necessarily use an online calculator for any OR analysis. Even if its a personal preference, I find that with a bit of effort and YouTube tutorials, one can easily grasp OR calculations depending on their requirements.

The online calculator may come in handy if the spreadsheet does not provide the feature you'd like to use. However, in most of my cases, it is kind of difficult to find an online calculator that fully satisfies my goal or objective. And also, most of the online calculators appear sketchy and unreliable. So, please be careful when using one.

O Solver do Excel é uma ferramenta de otimiza??o integrada ao Excel que pode ser usada para resolver problemas de programa??o linear, n?o linear e de otimiza??o inteira. Selecionando as variáveis a serem calculadas e as restri??es do or?amento, conseguimos chegar em valores ótimos individualmente ou coletivamente!

After years of experience working with data, I never touch Online calculators because we can do it with Spreadsheets. There are times where specific person which are beginners use these Online calculators, however when management wants to customized the calculation technique or the output it becomes bottle neck and they have to restart the work again in Spreadsheets. You may use those online calculators, however you may use it only for additional insight for yourself and leave all the analysis in Spreadsheets or other tools.

Online calculators are affordable tools offering various functionalities like financial calculations and statistical analysis. Many are free or low-cost, making them suitable for budget-conscious users. Some good options include Desmos for graphing, Wolfram Alpha for complex computations, and Google Sheets for collaborative spreadsheet work.

I agree the R and Python are the most prevalent and most versatile skill sets. I usually caution against making it an either/or decision. My experience is that it is best to know both. Even if you aren’t as proficient in one, the learning curve is less steep because of the underlying base principles.

Considering open source alternatives is certainly worthwhile when evaluating solvers for addressing operations research problems on a limited budget. The choice to adopt open source solvers depends on the scale and complexity of the problem at hand. While open source solvers can effectively tackle linear programming problems of reasonable size, the true test lies in their capacity to solve Mixed Integer Programming (MIP) and pure integer programming problems.

Open source software, such as Python and its packages PuLP, PythonMIP, Pyomo, OR Tools are good choices for implementing decision logic. There's also an excellent optimization framework for Julia language - JuMP. Combined with good open-source solvers like HiGHS, SCIP and CBC, these frameworks allow solving medium-sized optimization problems. For small problems and when quick prototyping is needed, GLPK - a free AMPL clone, is useful. AMPL community edition is not open-source, but free to use as well. All these tools are intended to work with a prepared input datasets, and can export solution to a database or a text file. So extra effort is needed to create some sort of user interface for data editing, scenario management and visualization.

R and Python are the most reliable ones when it comes to open-source software. By learning the language structure of either software, you may exploit it with almost no cost at all to assist you in completing your OR project. However, in my case, learning a programming language is not like flipping a hand. Rather, it is quite a challenging task to do. So, I suggest you to take it slow but consistent when learning R or Python language. Believe me, it will be worth it!

Open software are good, they are good and some level of community support is there. The choco library for java is my preferred tool for OS, and I also use the excel for the problem formulation. STEP 1: formulate the problem using excel, find constraints, variables and objectives functions. STEP 2: use the choco library to create the application to solve the problem formulated in step1. As good as your formulation is the solution achieved by choco will be optimum accordingly.

As analyst and a manager myself, I look into the background of the project, the scale of research and the objective first, only by getting the higher view of the project and how to solve it then you might choose the what tools should be used for Operations Research. I usually work with simple quadrant to make this decision - The scale of the research - Difficulties to get the research done After that I'm able to choose which tools to use, where most of the time I got back to Spreadsheets.

Choosing the best tool should be based on several factors surrounding the problem you are about to solve. This also depends on the problem type, the resource availability, data availability, skills availability, and many more. Understanding the problem is the first tool. For data-related problems, having tremendous and quality data is the bedrock for problem solving. However, this seems to be problematic most times for researchers. Most of a time, these data are not readily available. In summary, the best tools depend on the environment and nature of the problem

Choosing an Operations Research (OR) tool involves considering factors like problem complexity, ease of use, integration with existing systems, community support, scalability, cost, documentation, and performance. Assess the optimization problem's complexity, match the tool to your team's expertise, and ensure seamless integration. Active community support aids problem-solving. Consider scalability for evolving projects. Evaluate costs; open-source tools may be budget-friendly. Comprehensive documentation and resources are crucial for learning and troubleshooting. Finally, analyze the tool's performance in terms of solving time for your specific problems. Balancing these factors will guide you to the most suitable OR tool for your needs.

In my opinion, choosing the best tool to help you in an OR project depends on the complexity of the system or problem you are solving. When your problem is relatively simple, small, and straightforward, it will be better to use simple tools as well such as Microsoft Excel, Google Sheets, or Excel Solver. However, when you are dealing with a more complex problem that requires more sophisticated calculations but you are faced with a limited budget, then using an open-source programs such as R or Python may help you a lot. But as consequence, you need to at least understand how the programming language works for either program/software.

I agree with this and one of the main tools is the operations KPI indicators which helps you to assess where the strengths and weaknesses lay in the team. We had a KPI index created at one of the companies I worked and it was productive as we mentored the low performing members to do better and it worked.

If I had to pick two top tools for managing an operations research project on a budget, the first would undoubtedly be Python. With its extensive library ecosystem, including NumPy for numerical computation and SciPy for optimization, Python offers powerful, no-cost solutions for complex analytical tasks. Second, I’d recommend the open-source software Gurobi, which, through its academic program (yes, I’m cheating), provides free access to one of the most advanced optimization solvers available, ideal for linear programming, mixed-integer programming, and more. Together, these tools offer a robust framework for tackling a wide array of operations research challenges without breaking the bank.

Ultimately, OR algorithms are to be embedded in decision support solutions. For permanence of the solution, consider whether the open source tools being used will be well documented and supported in the future, and whether there is a strong community of users, so that you can find the talent when you need it in the future.

I would rather invest time and resources towards modifying data entry systems and improving data quality. Too often, the discussion is focused on what tools and solutions can be applied to process and analyze data in its current state. Most open-source (or lower-cost enterprise versions) tools will suffice. For example, I worked with a customer that wanted to develop an NLP application to read free text from maintenance logs and interpret a maintenance code (because the codes were often blank). The more obvious solution was to make the data entry system force the user to input a valid maintenance code and reduce as much other free text as possible.

The cost-effective ways of building OR solutions depend on the problem statement. There are a lot of open-source libraries or tools available which solve P problems optimally in realistic time. But they consume a huge amount of time to solve NP hard problems. In this scenario, one can strategically design an algorithm/heuristic/meta-heuristic approach to solve the problem at hand in a time efficient manner, that can give considerable (local optimal) results.

Absolutely agree! Spend your funds on data management and upkeep - garbage in, garbage out applies no matter what analysis tool you’re using. Data janitors should be invested in heavily!