FastPMRunner

Python code for running a suite of FastPM simulations. Prepare for running FastPM simulations in Latin hypercube for training emulators.

Installation for FastPM

FastPM GitHub website has explained several installation methods. Feel free to pick one you like.

For testing purpose, below uses the binary installation via Anaconda:

conda create -n cfastpm
conda activate cfastpm

conda install -c bccp cfastpm nbodykit

In the cfastpm environment, there should be a fastpm binary to run the simulation.

To run FastPMRunner, we also need matplotlib for plotting the output power spectrum (to see if the code works!)

conda install matplotlib

An example code for running 10 simulations with hubble parameter sampling in a range [0.65, 0.75] is provided in example/one_parameter.py:

python -c "from examples.one_parameter import *; one_parameter_hubble()"

After it's finished, you will find 10 simulation outputs will be saved in simulation_files/hubble_xxxx/ folder.

An example plot of how hubble parameter (h) changes the matter power spectrum is generated in simulation_files/pk_one_param_hubble.pdf: