[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: Systems for writing a science app front-end

Helen Faulkner wrote:
Neil Pilgrim wrote:
Interested to hear what people think wrt platforms/systems in which to
write a front-end for a console scientific app (typically long-running).
What would people recommend?

I use C++ + wxWidgets for the apps that run my (theoretical optics) simulations.
 I chose wxWidgets for writing the GUI mainly because it is available for
windows too, and this allowed me to cross-compile my code into windows .exes for
people in my group who don't use linux.

My current simulation code is also c++, but if I were to aim for multi-platform GUI, that wasn't too performance-sensitive, then I would be inclined to go for something like python/wxpython or similar. I'm sure that once C++ codes are compiled then they are easier to use/setup, but afaict python seems to something that (within reason) should just run with few changes - and installation can be managed through a .exe?

Not currently using windows, I wouldn't be keen to get involved in going through the (presumed) hassle necessary to get it compiled under windows - or do you use cygwin or msys(mingw) ?

I decided to switch to a GUI interface precisely to make it easier to see which
parameters were being changed when, and so I could easily do things like keep a
running count of the simulation results (I mainly deal with longish iterative
algorithms) and draw graphs representing the state of the system.  I'm sure
there are other things that would work as well, though, especially since qt is
now better licensed.

Do you use a wx plotting tool, or something like matplotlib? There seems to be some glue code to allow these to work together, which looks quite good.

Data management is certainly one thing that ends up costing me some time; in an ideal world a tool that could analyse my simulation input file and graphically represent the relationship between them would be incredibly useful - especially if you could look at an 'axis' of change (ie. changing a particular parameter over multiple simulations) and generate comparion plots. Of course, that would take some time to set up :)

I thought that QT under windows required a commercial license for distributing, or has this changed?


Reply to: