David, No problem at all! Thanks for the feedback. It is very useful. We'll see what happens with the simple survey, but up to this point each one of those five options had been a preference for 1 or 2 individuals. So there isn't or wasn't any overwhelming preference. Unless a strong consensus develops, I am thinking along much the same lines you indicate in your last two sentences:
- `diff(Eqn)` would raise an error (ValueError?) that would say the operation can only be applied to one side or the other with example syntax. The error would also direct the user to a special method (`ediff()`?). - `ediff(Eqn, wrt_list, const_list)` Eqn: the equation, wrt_list: symbols to differentiate wrt, const_list: symbols to hold constant. So this would be a fully specified partial derivative. - There is still the problem with how any derivatives in an equation would end up being evaluated. With SymPy's present capabilities there would need to be some autogeneration of functions to make this work with the differential equations module. - One stumbling block to bringing all of this together is that SymPy does not contain the concept of an infinitesimal of a symbol. I have not looked at it carefully, but some of the work necessary for this may be included in the differential geometry module. It is certainly embedded as an assumption in the integration module. Anyway. Thanks for the feedback. I will let the list know what the responses are. I'll probably wait a few more days to make sure people have a chance to respond. Jonathan On Saturday, February 6, 2021 at 11:16:23 AM UTC-6 da...@dbailey.co.uk wrote: > On 06/02/2021 03:07, gu...@uwosh.edu wrote: > > As the number of different people responding to this discussion has been > small I have put together the following simple 2 question survey about > expectations/preferences for differentiation of equations to try to broaden > the input. This is a simple google form. It collects google addresses only > to avoid multiple responses from the same google account. I expect it to > take people about 1 minute. The link is: > https://docs.google.com/forms/d/e/1FAIpQLSd-VCuOuNQtEo8VAuDA42SgNPYhMhr4Z5nEsGdwYoG9SCI5Nw/viewform?usp=sf_link > . > > I am sorry to be awkward, but I found this tricky to explain on the one > line provided for 'Other'. > > The definition of an equation seems so important (and remarkably tricky!), > would it be better to implement something that seems reasonable and put it > out as experimental code? > > Another possibility might be to simply fault any attempt to take the > derivative of an equation (and no doubt some other awkward operations) > until people had had time to experiment with equations for a bit. > > For example, if I chose option 5, I'd want to figure out how easy it would > be to set all the unwanted derivatives to zero, or stop them appearing. > > David > -- You received this message because you are subscribed to the Google Groups "sympy" group. To unsubscribe from this group and stop receiving emails from it, send an email to sympy+unsubscr...@googlegroups.com. To view this discussion on the web visit https://groups.google.com/d/msgid/sympy/c378757d-2207-4a99-9813-c7ed4efdd256n%40googlegroups.com.
