Skip to content

STEPBYSTEP_INSTALL.md

Latest commit

 

History

History
66 lines (50 loc) · 2.88 KB

STEPBYSTEP_INSTALL.md

File metadata and controls

66 lines (50 loc) · 2.88 KB

Step-by-step installation instruction

For Mac OS

  1. If you do not have a fortran compiler installed, install homebrew from https://brew.sh/
    Then in a terminal install with homebrew the required components:
    brew install gcc gfortran make cmake

If you did not do it yet, add homebrew binary directory in the shell variable PATH by adding IN THIS ORDER in .bashrc:

export PATH=/opt/homebrew/bin:/opt/homebrew/opt/make/libexec/gnubin:$PATH

(adapt it if needed). The order force to override the default binaries by the homebrew binaries and thus take gcc from homebrew and not the Mac OS preinstalled one.

  1. If you do not have python installed (the native python with Mac OS is generally outdated), install with
    brew install python

Installation of the main code and python package dependencies can be found at the end of the document, being OS independent.

For Windows

  1. If you do not have a virtual machine, install it from the app Microsoft store with the Window Subsystem for Linux.
    • Go to the store
    • Find Ubuntu, choose the distribution you like and install it (if you have some troubles for this part, have a look to specific helps here as example)
  2. Open a linux emulator terminal
  3. Follow the linux installation instruction here below

For Linux

  1. If you do not have a fortran compiler installed, install it with
    sudo apt install gfortran make cmake
    (or sudo apt-get install gfortran make cmake)

Main code and dependencies installation

This installation assumes that you have bash as shell.

  1. In the terminal, install the requested python packages, if not already installed, with
    pip install numpy scipy matplotlib ipython jupyter pandas getdist anesthetic
  2. Choose a directory and go on it
  3. Expand the zip file here or clone the program with git with
    git clone https://github.com/martinit18/nested_fit
  4. create the directory bin (mkdir bin) in your home (normally the place where you are when you open a new terminal)
  5. obsolete, see main README file for this part
    go to the decompressed directory and create the directory build:
    cd nested_fit-xxx (where xxx is the present version number)
    mkdir build
    cd build
  6. Go to the directory build and run
    cmake ..
    and then
    make
  7. In the file .bashrc in your home put the line
    export PATH=$HOME/bin:$PATH

Now if everything works, you can run nested_fit from terminal anywere you are.

Testing the code

  • Go to the distribution directory examples/data_analysis/gauss_bg
  • Run nested_fitxxx (where xxx is the present version number)
  • If everything works, the program will run and produce a series of nf_output_* files including nf_output_res.dat

More details on the program itself can be found in the README file.