Thanks Donaldson Tan, but this does not really answer my question. I'm thinking about evaluation complicated expressions involving huge sums of products of special functions taken at irrational points, where errors propagate, not just giving the n-th first digits of pi or of an already implemented standard special function
Best, Marc On Thu, Sep 25, 2025 at 5:35 AM Donaldson Tan <[email protected]> wrote: > > pi.evalf(5) gives you 3.1416 > > pi.evalf(10) gives you 3.141592654 > On Thursday, 25 September 2025 at 00:53:59 UTC+8 Marc Pegon wrote: > >> Hi everyone, >> >> First I'd like to thank the contributors to this project. >> >> I'm writing here because there is a question I couldn't find a clear >> answer neither on the online doc nor in the pdf doc concerning error >> propagation when using evalf. >> >> The doc says that that you can evaluate an expression to arbitrary >> precision by specifying n, increasing maxn if necessary, and setting >> strict=True. Even if the required precision is not reached, does evalf >> ensure that the printed digits are indeed the right digits in the decimal >> representation of the real number? Also, I wonder whether setting >> strict=True *guarantees* that the asked precision n is reached if >> PrecisionExhausted >> is not raised. If so, the next question is: how does evalf propagates >> errors? Does it involve interval arithmetic and how does that work with >> huge expressions involving a large sum of terms with products, irrational >> constants and special functions? >> >> Basically, I'm wondering if sympy can be used to prove/certify >> inequalities. >> >> I'm sorry if this question has already been asked and answered, but I >> couldn't find it precisely. >> >> Thanks! >> >> Marc >> > -- > You received this message because you are subscribed to a topic in the > Google Groups "sympy" group. > To unsubscribe from this topic, visit > https://groups.google.com/d/topic/sympy/MLvgspTMQ9U/unsubscribe. > To unsubscribe from this group and all its topics, send an email to > [email protected]. > To view this discussion visit > https://groups.google.com/d/msgid/sympy/78257cf4-cf34-4b8d-8e48-f6ce9ecd189fn%40googlegroups.com > <https://groups.google.com/d/msgid/sympy/78257cf4-cf34-4b8d-8e48-f6ce9ecd189fn%40googlegroups.com?utm_medium=email&utm_source=footer> > . > -- 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 [email protected]. To view this discussion visit https://groups.google.com/d/msgid/sympy/CAByOuZkvKAVQDyPCdCqeu2KYZXO45dLMhUPa-WEfn48oADiy3g%40mail.gmail.com.
