# Sampling with a distribution from the elements of a vector

28 views (last 30 days)
MiauMiau on 23 Jun 2017
Commented: MiauMiau on 23 Jun 2017
Hi,
Say I have an 1xm vector with any kind of numeric entries (for instance doubles or integers). As an example, I would have the vector a:
a= [0.1, 0.3, 0.5, 0.55,0.6]
Now I would like to sample from this vector with a distribution. Say I would want to use a normal distribution, then the mean would be basically the middle element of vector a. That would imply that "0.5" would be sampled most often, then "0.3" and "0.55" second most, and "0.6" and "0.1" the least number of times. How could I do that? In my concrete case, I would want to use left-skewed distribution. Any help appreciated.

the cyclist on 23 Jun 2017
Do you have the Statistics and Machine Learning Toolbox? If so, you could use the randsample function. Specifically, you would use the input parameter w for the weighting appropriate for the distribution you want to sample.
MiauMiau on 23 Jun 2017
Ok, will have a look at it and write questions here if I will have any. Thank you!