hi,

does anyone know what are some of the better curve-fitting and optimization programs out there? are common programs like Origin and gnuplot good? how about pricy ones like Mathematica and Matlab? Or general software like Excel?

I have been using Excel for optimization purposes for a while, but I notice that sometimes the fits are not very good...and sometimes the program even reports dubious equations for graphs. i am guessing that there are some bugs in there because Excel is not dedicated for such intensive mathematical analysis in the first place. i have been introduced to gnuplot bys some professor, but before i go on to waste days getting used to the command lines, i would like to get some opinion.

Hi Limlin

Not sure what your field is, so this may not be especially useful

I'm doing pharmacology, and find that for standard curves, and general curve fitting, a programme called graphpad prism is just the job.

I think it is designed specifically for the biological sciences, so you may find it to be of limited use for what you want, but it does have many options that could come in useful. It's way better than Excel, and a little more user friendly in my opinion.

Anyway, you can download a month's free trial, google it, and see what you think

Personally, i think getting to know Matlab is time well spent as it's a useful line for your CV. Can you acquire it from someone in the department?

hmm, i think Matlab is not that easy to get, because the license is an issue. furthermore, its a huge software, so if i use it only for drawing curves, it might not go down well with my department.

i am doing engineering work. will go take a look at graphpad thx =))

actually i find excel really useful, in no small part because it presents your data in a form that is very easy to read and manipulate, and there are many hidden functions that i am just discovering.

Hi Limlim: MATLAB is definitely excellent and powerful. I have used it a fair amount for curve fitting and optimization. It's straightforward to import data from Excel, graph, etc. The only challenge would be a little bit of programming...but it's not too difficult if you have some background in programming. MATLAB offers a lot of flexibility and power at the cost of a little more effort setting things up. MATLAB is downloadable....not that I encourage that, of course! Also, I would have thought you'd have access to it in computer labs. At my university, it's as common as Word.

hi bald_monkey, i don't have access to Matlab at where i work, because we already have idl, so the argument is that idl is as good as Matlab. i think they are right, just that i have no knowledge of idl at this point in time. have you ever compared your optimization results with other programs, like gnuplot or excel? and where can Matlab be downloaded?

Hi Limlim,

I do not know what field or application you have but if it is for elisa analysis, MasterPlex ReaderFit will be a good fit (no pun intended). It has a Best Fit feature which optimizes based on 3 parameters that you can choose from: root mean square error, r-square, or standard deviation of percent recovery. It is extremely user friendly and you will find the learning curve to be minimal. There is a fully-functional 14-day free trial and a short demo movie if you are interested:

http://www.miraibio.com/masterplex-readerfit/curve-fitting-for-plate-readers.html