Skip to content
This repository has been archived by the owner on Feb 22, 2024. It is now read-only.
/ MLplayground Public archive

Web application for collaborating to build ML models for Tracker DQM/DC

Notifications You must be signed in to change notification settings

CMSTrackerDPG/MLplayground

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

DQM Playground

Django CI - Updated Code style: black

The goal of the DQM Playground is to serve information from various sources (OMS, Run Registry, DQM GUI, static files from the ML4DQM effort) in order to ease model development and to provide a place to compare the predictions of the various models.

Requirements

  • A postgresql database.
  • A ROOT installation.
  • An .env file which contains the variables listed below.

Environmental variables

All must be stored in a file named .env:

DJANGO_DATABASE_ENGINE
DJANGO_DEBUG
DJANGO_DATABASE_NAME
DJANGO_DATABASE_PASSWORD
DJANGO_DATABASE_USER
DJANGO_DATABASE_HOST
DJANGO_DATABASE_PORT
DJANGO_SECRET_KEY
DIR_PATH_DQMIO_STORAGE

DIR_PATH_DQMIO_STORAGE (str)

List of directories to look into for DQM files (csv, nanoDQMIO). Separate multiple directories with a colon (:), e.g.:

DIR_PATH_DQMIO_STORAGE = "/dir/to/dqm/files:/this/is/another/dir"

Behavior

Management Commands

histograms

  • exctract_lumisections_histos1D_csv: Given a CSV containing 1D Lumisection Histograms, this command will parse the file's contents and create appropriate entries in the LumisectionHistogram1D table.
  • exctract_lumisections_histos2D_csv: Given a CSV containing 2D Lumisection Histograms, this command will parse the file's contents and create appropriate entries in the LumisectionHistogram2D table.

Authentication

External applications

To query the API from a third-party application, you will need to generate a token (from the admin dashboard)

Development

Database

For storing histograms, the ArrayField is used, so a PostgreSQL database is recommended for running the project locally.

Installing pgadmin4 is also recommended for easier interfacing with the DB.

API access via JS

Accessing the API via a rendered HTML page requires session authentication of the DRF. This is done by:

  1. Requiring the user to login to the page that contains the JS that does the requests
  2. Configuring DRF to accept SessionAuthentication
  3. Adding the X-CSRF-TOKEN header to the API request and setting its value to the CSRF token which you can get by accessing the cookies stored in the current page. [https://docs.djangoproject.com/en/4.0/ref/csrf/#ajax]

[https://www.django-rest-framework.org/topics/ajax-csrf-cors/]

Building the Dockerfile locally

If outside the CERN GPN, you will need to create a bridge (via, e.g. sshuttle) first.

Then, cd into the directory where the Dockerfile is and run:

sudo docker build . --build-arg ROOT_TAG_NAME=v6-24-08 --network="host"

About

Web application for collaborating to build ML models for Tracker DQM/DC

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Contributors 4

  •  
  •  
  •  
  •