RAMP starting kit on Primary Vertex reconstruction

Set up

1. clone this repository

git clone https://github.com/ramp-kits/vertex_finding.git
cd vertex_finding

2. install the dependancies

• with conda

conda update conda                   # make sure conda is up-to-date
conda env create -f environment.yml  # use environment.yml to create the 'vertex_finding' env
source activate vertex_finding       # activates the virtual env

• without conda (best to use a virtual environment)

python -m pip install -r requirements.txt

3. download the data

python download_data.py

After download, the data will be unpacked to data/train and data/test (might take a while). By default, this will extract 5k train and 1k test events.

Building the baseline solution (using C++)

The baseline solution was adapted from the Primary Vertex reconstruction used in LHCb.
It requires CMake, the Boost library (both installed with conda) and a decent C++ compiler to be built.

• manual build

  mkdir build && cd build
  cmake .. && make
  cd ..

• automated build, when calling the baseline submission for the first time

  ramp_test_submission --quick-test --submission baseline

Testing the kit

After setting up the environment, run the starting kit (random values) and the baseline solutions.

ramp_test_submission --quick-test
ramp_test_submission --quick-test --submission baseline

They should run on a subset of the data and print out the scores.

To process all of the locally available data, remove the --quick-test flag.

ramp_test_submission --submission baseline

Local notebook

Get started on this RAMP with the dedicated notebook.

Help

Go to the ramp-workflow wiki for more help on the RAMP ecosystem.