Box-Muller Transform + R Demo

[@shankaradevapriya153]

just what i needed for tomorrow.. thanks and keep it up..

[@tim-duncan2137]

thanks professor well done

[@lg_susdiaz]

Thank you for your explanation. I was reading in a lot of pages and books but I couldn't understand how to get them, but you explained very well, you save me.

[@taotaotan5671]

As a side, I learned that in early days people average 12 uniform distribution to approximate normal distribution. It clearly based on central limit theorem, but still found it amazing.

[@prasannakumarkottakota68]

Thank you, Nice explanation ..

[@algorithmo134]

we need more of this! stats video and probability theory!

[@Gengar99]

Nice, thank you so much.

[@fa-pm5dr]

i used box muller in a game cause that back, Godot didnt have a normal distribution sampler, only randf

[@SagnikMajumder-rc6ct]

Nice video, thank you. I just had one query that why we do not use the uniform random uniform U_1 also for getting the theta values ? Like why we do not write theta = 2 * pi * U_1 rather than using U_2 ? Is there any problem arising if we use that ?

[@nathanroe]

can box-muller be used to sample from normal distributions with different means and standard deviations?

[@amirhosseintalebi6770]

thanl you. But, Box-Muller Transform reference not found

[@mathetal]

keyonvafa.com/box-muller-transform/

[@nathanroe]

why does theta equal 2pi U2?

[@weiyang145]

Why don't we just use code "rnorm(1)" to generate a standard normal random variate? Can i assume the result what will i get by using rnorm(1) is the same as Box-Muller transformation. Hope u can solve my doubt, tq very much!

[@mathetal]

good question! rnorm does return random samples from a normal distribution. So if you think about it, "rnorm" is just a function, and there has to be some underlying algorithm that is called every time someone calls that function. You can check out the "random number generation" page in the R documentation for more information, but actually R has several algorithm options to use for RNG and Box-Muller is one of those algorithms, although it is actually not the default choice for rnorm.

[@yulinliu850]

👍