Monte Carlo Simulation , a simpler version

[mohsenhs82]

Hi all,

I have asked this question before, but it seems I was not asked it in an appropriate way, so I draw the following flowchart which I think explain the problem much simpler and faster. I have the following situation and would like to do a Monte Carlo Simulation for this. Could you please kindly let me know how I can do this?

Any help is highly appreciated.

Many thanks
Mohsen

 

[rogojel]

hi,
I would be a bit skeptical about the differences in the shape of the distribution - that could easily be a sampling accident. Maybe you can approximate all the samples sufficiently well with a Weibull?
regards

 

[mohsenhs82]

Hi Rogojel,

Thanks for your consideration. It might be the sampling problem or insufficient sampling data, however I do not have more data. This data set is actually the arrival rate in 24 hours. For some hours fitdistrplus showed Weibull as the best and for some others Lognormal was the best (by considering AIC). I am trying to implement a Monte Carlo simulation and test different rate of sampling (i.e. 0.05,0.1,0.15 to 1) and investigate how my approach/algorithm is biased. The above flowchart is the simplified version of what I did so far.

Thanks once again