I was wondering if anyone knew of good approximations of the primorial function for large numbers, or of reasonable bounds for it. By primorial I mean:(adsbygoogle = window.adsbygoogle || []).push({});

[tex]n\#=\prod_{p\le n}p[/tex] forpprime.

All I know is [tex]n\#\sim e^n[/tex] and the trivial [tex]\pi(n)!\le n\#\le n![/tex].

For small numbers (n < 1,000,000,000), e^n is larger than n#. Does this hold generally?

Obviously, the crude upper limit can be hacked slightly: take out the evens from the factorial and multiply by 2, take out the multiples of 3 and divide by 3, etc. In the extreme case this is just calculating the primorial itself, so not very useful. The lower bound can be improved to

[tex]p_n\#\le n!\prod^n_{k=5}\ln k+\ln\ln k-1[/tex]

using lower bounds on the primes, but this is unwieldy, and I'm pretty sure [tex]\lim_{n\rightarrow\infty}\frac{p_n\#}{n!\Pi}=0[/tex] where the product is as above. Using the best bounds out there (by Dusart) improves the result slightly at the cost of complicating it further.

**Physics Forums | Science Articles, Homework Help, Discussion**

Join Physics Forums Today!

The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

# Primorial: asymptotics and bounds

**Physics Forums | Science Articles, Homework Help, Discussion**