I’m performing 10 independent Monte Carlo simulations, each of them gives me a histogram. Then I ‘hadd’ these 10 histograms in order to have a single total histogram (let’s call it Htot).
With ‘hadd’ I automatically have the mean and the standard deviation of that Ht and now I’m wondering how I could have the error on the standard deviation of Ht ? In order to have error bars on the StdDev.
Moreover how the ‘hadd’ command is calculating the new standard dev and the new mean from the 10 aforementioned histograms. Is it simply a sum of each StdDev of each histograms ? And a sum of the means divided by 10 ?
For now I’ve calculated by hand the error as the fourth moment but it’s very long and I’m not sure anymore of what I’m writing down… anyway I hope there is an “automatic” way of doing that.
I hope I’m concise enough, if not I’ll give more details.