Normal/Gaussian Function or Method: Normal Probability Density

#1 Опубликовано: 06.01.2023 19:08:49

4 сообщений из 90 понравились пользователям.

Группа: User

This matrix generator (thanks gUrrozEN) produces random 'equal-probability' values for a given center value and tolerance band.

I'm hoping to revise this to return values within a normal/Gaussian probability as illustrated:

Random to Normal.sm (42 КиБ) скачан 21 раз(а).

Any thoughts?

#2 Опубликовано: 06.01.2023 22:47:00

Jean Giraud

983 сообщений из 6866 понравились пользователям.

Группа: User

Wrote
Any thoughts ?

Native Smath plugins generate uniform randoms.
The Central Theorem proves that enough uniform randoms
generate a Normal Distribution.
You simply need to bin in an histogram utility.
Is that is what you are looking for ?
Otherwise, I don't understand your request.
Cheers ... Jean.

#3 Опубликовано: 07.01.2023 02:07:39

Mark R Harris

4 сообщений из 90 понравились пользователям.

Группа: User

Thank you for the reply Jean.

I'm after a method or function that will return a random value with the probability that 68.2% of the time it will be within 1 sigma, and 95.4% of the time it will be within 2 sigmas, etc.

#4 Опубликовано: 07.01.2023 04:13:06

Alvaro Diaz Falconi

992 сообщений из 1674 понравились пользователям.

Группа: User

Hi. For 'convert' random uniform distributed numbers between 0 and 1 to another probability distribution, in your case, for get some normal distributed numbers you can use the inverses of the cdf distributions. You can check that here for 1000 numbers you have the desired mean and std deviation

so, I guess that in your function rmat this is the function for normal distributed numbers

random to nomral.sm (5 КиБ) скачан 29 раз(а).

Best regards.
Alvaro.

1 пользователям понравился этот пост

Mark R Harris 07.01.2023 05:25:00

#5 Опубликовано: 07.01.2023 05:32:57

Mark R Harris

4 сообщений из 90 понравились пользователям.

Группа: User

Thanks Alvaro, very greatly appreciated.

For me, the sample file has a problem:

#6 Опубликовано: 07.01.2023 05:55:38

Alvaro Diaz Falconi

992 сообщений из 1674 понравились пользователям.

Группа: User

Hi. I mean: substitute that M[i,j] into your rmat(...) function for get normal distributed numbers with the rmat algorithm instad uniform distributed ones.

Best regards.
Alvaro.

1 пользователям понравился этот пост

Mark R Harris 07.01.2023 06:22:00

#7 Опубликовано: 07.01.2023 11:22:38

Jean Giraud

983 сообщений из 6866 понравились пользователям.

Группа: User

Wrote
Thank you for the reply Jean.

I'm after a method or function that will return a random value with the probability that 68.2% of the time it will be within 1 sigma, and 95.4% of the time it will be within 2 sigmas, etc.

I don't understand completely ... sorry.
However, from a native random sigma, you can solve for a desired other one.
Cheers, be good Down Under ... Jean.