Skip to content

Commit

Permalink
Minor updates to the documentation
Browse files Browse the repository at this point in the history
  • Loading branch information
honeygupta committed Aug 20, 2020
1 parent cfccd3f commit a83999e
Showing 1 changed file with 4 additions and 4 deletions.
8 changes: 4 additions & 4 deletions README.md
Original file line number Diff line number Diff line change
Expand Up @@ -115,7 +115,7 @@ A summry of the experiments performed and the results obtained as a part of [Go

4. `postprocessing.py`: holds various functions for saving data back into the ROOT fileformat.

* ### Process_data
* ### process_data
This repository contains python scripts that can be used to create the training and testing datasets from ATLAS, PhenoML and DarkMachines datasets.

All the python script have very similar codes and functions, except for some small variations depending on the experiments for which they were used. The names of the python files should be self explanatory to define the task and the kind of dataset they create.
Expand All @@ -124,19 +124,19 @@ A summry of the experiments performed and the results obtained as a part of [Go
1. `plot_particle_distribution.ipynb`: contains the functions to plot the particle distribution for a particular process (from the PhenoML dataset). It also contains the scripts to create data distribution plots for different process (.csv) files.
2. `process_data_as_4D.ipynb`: gives a visual intuition about different parts of the processing scripts and their functions.

* ### Scale data all
* ### scale_data_all
This folder contains a script that scales (or normalizes) the data generated by the `processing` scripts. The script uses `FunctionScaler` scaler to normalize the data. This was used in the experiments mentioned in Eric Wulff's thesis.

However, during our experiments, we shifted to standard normalization and mainly to custom normalization, which is implemented as a part of the training and testing scripts.

_Keeping this script for the sake of completeness._

* ### Train_models
* ### train_models
Throughout the project, the experiments were run using a batch service at CERN called HTCondor. This folder contains the scripts that were used for submitting different training jobs during the experiments.

Have included these to ensure reproducability and for making Knowledge Transfer easier.

* ### Examples
* ### examples
This is the folder that contains the training, testing and analysis scripts for all the experiments for the abovementioned three datasets, with standard and custom normalization.
1. `phenoML`

Expand Down

0 comments on commit a83999e

Please sign in to comment.