Inverse Transform Sampling | Triangular Distribution

[@mahmoudel-khalil3328]

Thank you, you saved my life. Your videos are awesome.

[@robertwright4906]

Amazing work!

[@martinsanchez-hw4fi]

In 3:11 there is a problem. You say that a CDF must be >=0, but when you invert a CDF, you are not getting a CDF, you are getting a random variable, which can be (and in the general case will be) also negative. For example, if I have a normal distribution, its quantiles are from -infinity to +infinity. In this case it counfuses me when to take the infimum (you do it in the second case, but not in the first). Even though its correct, again, the argument that the inverse must return positive values is not right.

[@MisterRay15]

What do I do if I have 3 unknown points for the triangle, minimum, maximum and mode?

[@RafiqulIslam-dy8cm]

Could you please give an example of order statistics

[@syh1414]

Could you show the R coding at the end? Thank you!

[@JoseVinicius-fe3vk]

it is supposed to be 1/2 + 2x - x^2/2 -1

[@hashbrowncookie8444]

Just like the other comment said, can you please give a rough idea of how you would be writing this using R? Please.

[@hashbrowncookie8444]

I am not sure how you would include the limits of y in your code? I am having difficulties figuring out that part.

[@samalkayedktaishat9927]

hello , how do i do this using R ??

[@muhannadobeidat]

Here is my solution for R: Define function for the new CDF. Pass array of uniform distribution output from 0,1 for 1000 samples, apply the CDF for each input and plot output. calculateY

[@farkas5572]

@@muhannadobeidat Shouldn't it be x

[@alf2954]

@@muhannadobeidat Bro, you've saved one hour of my life

[@뚜둥-b8e]

코딩 안하고 딱 끝나버린 이유는 뭐냐 ?