Statistics and Data Analysis (HEP at LHC)
Prof. dr hab. Elżbieta Richter-Wąs
Slides extracted from W. Verkerke lectures at SOS School 2018, France
Coding probability models and likelihood functions
2
RooFit, RooStats and HistFactory
3
RooFit core design philosophy
4
RooFit core design philosophy - Workspace
5
Basics – Creating and plotting a Gaussian p.d.f.
6
Basics – Generating toy MC events
7
Basics – ML fit of p.d.f. to unbinned data
8
RooFit code design philosophy - Workspace
9
The workspace
10
Using a workspace
11
Accessing a workspace contents
12
RooFit core design philosophy - Workspace
13
Factory and Workspace
14
Populating a workspace the easy way – „the factory”
15
Model building – (Re)using standard components
16
Model building – (Re)using standard components
17
The power of pdf as building blocks – Advanced algorithms
18
The power of pdf as building blocks – adaptability
19
The power of pdf as building blocks – operator expressions
20
Powerful operators – Morphing interpolation
21
Powerful operators – Fourier convolution
22
Example 1: counting expt
23
Example 2: unbinned L with syst.
24
Example 3: binned L with syst.
25
Example 4: Beeston-Barlow light
26
Example 5: BB-lite and morphing
27
HistFactory – structured building of binned template models
28
HistFactory elements of a channel
29
HistFactory elements of measurement
30
Example of model building with HistFactory
31
Example of model building with HistFactory
32
HistFactory model output
33
HistFactory model structure
34