Skip to content

Latest commit

 

History

History
74 lines (42 loc) · 3.4 KB

Histo_readme.md

File metadata and controls

74 lines (42 loc) · 3.4 KB
Raw

Histogramming Guide

This guide shows how to fill histograms with event data sklimmed in the preprocessing phase.

Filling Types

There are 2 types of filling availalbe to users: simple fill and cross product fill with multiple samples, regions and systematics.

Running Simple Fill

To run a simple fill change directory to pythium folder and run python3 scripts/histogramming.py. This will load event data, fill histograms and save them in a directory specified in histogramming_config.py.

Configuring Simple Fill

The configuration file histogramming_config.py contains important parameters that control filling process. The following variables affect the simple fill:

  • client_params dictionary of arguments passed to the Dask client.
  • file_list dictionary of arguments that controls the function that creates a list of files to be used in the filling process.
  • computation_params dictionary of arguments that controls which observable dictionary is used in the filling process and how many files are processed in each batch.
  • out_dir specifies output directory.

Running Cross Product Fill

To run a cross prodcut fill change directory to pythium folder and run python3 scripts/cross_product.py. This will load event data, fill histograms in a cross product manner and save combinations to a directory specified in histogramming_config.py.

Configuring Cross Product Fill

The same configuration file histogramming_config.py contains other parameters that describe the cross porduct:

  • Samples_XP contains a list of samples that are used in the cross product.
  • Regions variable is a list of regions that are used in the cross product.
  • Systematicscontains, as above, a list of systematics used in the cross product.

These lists contain 3 objects that are important for the cross product performance, these are respectively: XP_Sample. XP_Region and XP_Systematics.

Sample Object

XP_Sample is an object that contains list of files for each sample. Upon initialization, file list is compiled. This can be done in 3 ways, user can pass a list of files directly.

XP_Sample(name = 'test', file_list = ['tmp/data/file.h5'])

User can also pass a top directory to start a search for files and a regex filter.

XP_Sample(name = 'test', regex = True, top_directory = 'tmp/data/', file_regex = '(.*file.*)(.*h5$)')

One can also pass Sample object from the preprocessing stage.

XP_Sample(Sklim_Sample = item, top_directory = 'tmp/data/')

Region Object

XP_Region is an object that contains different regions and their selection criteria. During cross product filling, these selection criterai are applied to the data set.

XP_Region(name = 'test', filter = 'col1 >= 1')

Filter argument uses pandas query function.

Systematic Object

XP_Systematics is a parent object of different kinds of systematics:

  • XP_Overall changes the weight of the entire histogram by a set amount.
  • XP_Formula changes based on a function defined by a user.
  • XP_Histo subtracts specified histogram from one another.

Systematic Object Examples

synatx explanation
XP_Overall(name = 'test1', Adjustment = 0.5) Increase each weight by 50%
XP_Formula(name = 'test2', Formula = myfunc) Change weights according to the formula provided in myfunc function
XP_Histo(name = 'test3', name_of_hist = hist_object) Subtract hist_object from a histogram