Analysing samples and plotting

Roofit and RooStats
1

I have an ascii file with huge number of lines of data from a monte carlo run. The lines are in the form

x1 x2 x3 x4 … probabily likelihood

Any ideas/help on how to make 1D 2D and 3D maginalised plots, with 68% and 98% confidence level contours, of these using ROOT will be greatly appreciated

Cheers

2

Hi,

if you want to be helped, could you please elaborate a bit more on what your data exactly are and on which plots would you like to make, maybe giving some examples. Thank you,

Lorenzo

3

the data file have 6 columns say. the first 4 are the parameters (x1 x2 x3 x4) of the problem and their corresponding bayesian posterior probability and likelihood are on the 5th and 6th column. I want a 2D coloured histogram of say x2 and x3 marginalising the other parameters. I have attached an example data file. Hope this clarifies the request/question. Thanks
file.txt (15.7 KB)

4

Hi,

I don’t think we have yet in ROOT the bayesian statistical tools for producing these plots. We have the possibilities to produce profile likelihood plots (with frequentist statistics), showing for example 68 and 96 confidence levels contours using Minuit.
Best Regards

Lorenzo