I'm looking for a good cross platform (mac, windows & linux) python GUI framework / library that will make my life easier while writing a data analysis program. Since my data is represented by custom data classes, it would be great if the GUI framework / library could take away the burden of having to code input checks, validation, etc (i.e., create input dialogs that take care of checking for the correct data range / data type based on the data model). The only library that I've found so far is TraitsUI. Are there more (similar) libraries / frameworks out there?
Then Enthought Suite (not just TraitsUI) is the most complete as it provides everything from building the model to showing it including input validation. It plays nicely with numpy and scipy which is nice for a scientific app. Enthought UI can use Qt (via PySide or PyQt) or wx as backends.
You can also use Qt directly via PyQt or Pyside and embed plots using matplotlib or PyQwt. QtDesigner allows you to generate nice UI with very little effort. You can achieve the type of initialization, validation as with Traits but with more effort.