Slang (1960s, NASA), PROSE (1970-1980s, Time Sharing computers, see PROSE-development-history-Wikipedia article), & FortranCalculus (1990-today, PCs w/Windows, see DOI: 10.13140 / RG.2.2.31951.18080) languages give the users the option of tweak parameters as one solves their math problem(s). Are there other languages available today that will allow users to tweak parameters? Languages must be able to create executable (EXE) files. Please state their names and any links that may be helpful.
Could you give an example of tweaking parameters? You have some numerical method which doesn't give the answer you want and you want to run an algorithm again with different parameters? I have worked in Operational Research departments where they'd routinely run, in parallel, models and graph the models to determine how the outputs are affected by parameters. My own research interests are in natural language (I program in NL using Enguage language engine which allows you to call other programs) and it would be interesting to see how you determine what constitutes a tweak. Or are your problems solely written? It would be interesting to see if it were possible to run an algorithm implementation in parallel by voice command, and determine the output goals by voice: "run myprog with time t varying from 5 to 20 seconds in 2 second intervals". Need to interrogate results at the same time? "where does the maximal throughput r occur" ? It could be that tweaking parameters sounds simple, but is complex?
Martin,
Many people have equations that require the user to find to best parameter values for their problem. Curve Fitting may be the easiest to visualize. The user has their equations y = f(x; a, b, c, etc.) where a, b, c, etc are the parameters that need to be found for the best curve fit.
All or most Optimization problems tweak parameters in order to find the optimal solution. The Oil Refinery Production ... see https://www.researchgate.net/publication/326812480_Oil_Refinery_Production_What_is_the_company's_goal_Maximize_Profit_Minimize_Pollution_Other ... is a great example too. Here an oil corporation builds models of their refinery processes, finds models for their constraints e.g. Inventory, source of crude oils, etc. Then they try to simulate what each of their refineries produces daily. Once they have duplicated their present day production, they move on to optimization where the computer tweaks hundreds of parameters in order to meet their objective, e.g. maximize profit while minimizing pollution! These optimization runs may take days or weeks to run to completion. With new computer designs these runs should be completed in 2 hours, we hope.
As for complexity, users are just concerned about their math models (i.e. Equations & Constraints). The oil example has thousands of Partial Differential Equations (PDEs) to solve while finding hundreds of parameters. But all that is left to the solvers (i.e. numerical methods). The user may try several solvers to convince them that the optimal solution is truly optimal. :)
FortranCalculus (tm) is the language I use and recommend for any modeling, simulation, and optimization problems ... see https://www.researchgate.net/publication/326624021_FortranCalculus_Calculus-level_Language_Problem_Solving_Presentation ... its freeware! Try it, you may like it as I do.
Phil,
I'm afraid I'm going to have to step down from this challenge, interesting as it looks. I think you goals are too far from the abilities of my software at the moment, as you are in a deep mathematical sphere. But thanks for the clarification and good luck in finding collaborators!
Martin
NASA's Apollo Space program funded the Slang language that got us to the moon. This is a by-product of Slang (1960s) for Windows.
- Badges
- Science topic
- Similar topics
- Computer Science and Engineering
- Programming Languages