0
$\begingroup$

I have to program this in some environment so I won't be able to access other softwares.

Let's say I've got some 50 numbers that is more or less skewed to one side. If we are to assume that it is like a normal distribution, there's a handy equation that simply calculates the standard deviation, and that would already be the 34.1% interval from the median.

I would like to calculate the interval from the median of the skew normal distribution model fitted to my data, 25% or 34.1% it doesn't matter, just something similar. How exactly should I program to calculate that given my set of data? I am basically trying to get an asymmetrical interval out of this.

Thanks in advance.

Improve this question
$\endgroup$
1

1 Answer 1

Reset to default
-2
$\begingroup$

you can not consider skewed dataset to be normal. You should make it Normal for further stat. investigations (if there nature is really Normal in theory) - to have Normal Distribution of, e.g., X's logarithms & deal this distribution as Normal further on... here are some other ways to deal with skewed data (e.g. in ML overall, not only academical stat.). However, most commonly used method in stat. is LOG(feature), just don't forget to make reverse transformation to 'feature'(as is X) at the end of whatever stat. calculations to get initial vars of this X (so-called 'un-log' at the end)...

Improve this answer
$\endgroup$
2
  • $\begingroup$ As it’s currently written, your answer is unclear. Please edit to add additional details that will help others understand how this addresses the question asked. You can find more information on how to write good answers in the help center. $\endgroup$
    – Community Bot
    Commented May 29, 2022 at 7:28
  • $\begingroup$ I've just answered about the subject of the Q, that tells "Let's say I've got some 50 numbers that is more or less skewed to one side. If we are to assume that it is like a normal distribution" -- my answer: you cannot assume, you should transform skewed data to normal distribution (e.g. with LN) -- in order you could get trustfull results of stat. calc. further on... If you don't need Normal stat - I don't know about what % the question could be about in skewed data. - it looses sence. I advise to learn if you're dealing correctly with your skewed data before asking how to do it incorrectly. $\endgroup$
    – JeeyCi
    Commented May 30, 2022 at 13:50

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.