On Wednesday, March 28, 2012 7:53:53 PM UTC+5:30, rl wrote: > > Hi, > > > > You should also start writing up your application, so you can get > > > some feedback on it. Something to think about: you have a lot of > > > ideas and you may not be able to get to all of them over the > > > summer. This is fine, but you should think about prioritizing them > > > so that you do the most important things first, and put the less > > > important things to the end of the summer for if you have time. > > > Consider not only the relative importance of each idea, but also > > > which things will depend on one another (e.g., will you need to fix > > > up the structure of the new plotting module before implementing > > > this algorithm, or can they be done independently?). > > > > > > I have started working on my application. I think it will be ready by > > Friday. > > I am thinking about implementing the svgfig plotting backend at the > > end of summer, if I have time. > > I want to purely concentrate on interval arithmetic and the implicit > > plotting > > module. I won't be fixing the structure of the new plotting module as > > they are not at all dependent of each other. The primary goal of my > > plotting module will be the ability to plot implicit functions, while > > the plotting module by Krastanov will handle functions of the type > > y = f(x) . So I think they will be completely independent. > > No they are not independent, at least not if you use the algorithms in the > paper: > "Reliable Two-Dimensional Graphing Methods for Mathematical Formulae with > Two Free Variables". > > If you can plot f(x,y) = 0 then you get the case y = f(x) for free! >
Yeah right. But I am guessing the plotting time will be large for the algorithm in the paper to plot y = f(x) which we can do much faster by sampling. > > I would like to know whether I can edit mpmath codebase in sympy. > > This is necessary as mpmath interval arithmetic library does not > > support all the functions. > > These additions should better find their way upstream into the > mpmath sources. But I assume you could do this later. > > > Also I would like to extend their module to allow property checking. > > What do you think will be better? > > Is property checking a worthwhile goal in general interval arithmetic > or is it just a helper tool for plotting? > It is just a helper tool for plotting. As mentioned in the paper, we would like to handle plots like y < sqrt(x). mpmath gives a result something like below for sqrt([-0.5,0.5]) In [10]: a = iv.mpf([-0.5,0.5]) In [11]: a**0.5 Out[11]: ([-0.70710678118654757274, 0.70710678118654757274] + [-0.70710678118654757274, 0.70710678118654757274]*j) In [12]: b = iv.mpf([-5,-3]) In [13]: b < a Out[13]: True So I will have a wrong result. So property checking will be only for plotting. Thanks, Bharath M R > -- You received this message because you are subscribed to the Google Groups "sympy" group. To view this discussion on the web visit https://groups.google.com/d/msg/sympy/-/A_Tb0pMHlLAJ. To post to this group, send email to [email protected]. To unsubscribe from this group, send email to [email protected]. For more options, visit this group at http://groups.google.com/group/sympy?hl=en.
