Running iDEA

As it is a python package there are many different ways of running iDEA.

Using the iDEA code directly

Use idea-run to produce the default parameters file:

idea-run

In order not to overwrite results from different calculations, make sure to choose different run names for different inputs

### Library imports
from __future__ import division
from iDEA.input import InputSection, SystemSection
import numpy as np


### Run parameters
run = InputSection()
run.name = 'run_name'                #: Name to identify run. Note: Do not use spaces or any special characters (.~[]{}<>?/\)
run.time_dependence = False          #: Run time-dependent calculation
run.verbosity = 'default'            #: Output verbosity ('low', 'default', 'high')
run.save = True                      #: Save results to disk when they are generated
run.module = 'iDEA'                  #: Specify alternative folder (in this directory) containing modified iDEA module
run.NON = True                       #: Run Non-Interacting approximation
run.LDA = False                      #: Run LDA approximation
run.HF = False                       #: Run Hartree-Fock approximation
run.HYB = False                      #: Run Hybrid (HF-LDA) calculation
run.EXT = True                       #: Run Exact Many-Body calculation

Using the iDEA package in a python script

Since iDEA is designed as a python package, it can be run from everywhere, if you let your python installation know where the package is located. During the installation of iDEA the idea-public directory should have been added to PYTHONPATH. To test this has worked simply perform the following

cd $test_folder                  # some folder you have created
cp $path_to_iDEA/iDEA/parameters.py . # where you have downloaded iDEA
idea-run

Here, we are running iDEA much in the same way as before but your $test_folder can be located anywhere on your computer.

The main advantage of having an iDEA python package is that you can access its functionality directly in a python script.

The example below uses a python loop to converge the grid spacing for an iDEA calculation of a test system of non-interacting electrons in a harmonic well.

from iDEA.input import Input

# read parameters file
inp = Input.from_python_file('parameters.py')
inp.run.verbosity = 'low'

# Converging xmax parameter
for xmax in [4,6,8,10]:
    # Note: the dependent sys.deltax is automatically updated
    inp.sys.xmax = xmax

    # perform checks on input parameters
    inp.check()
    inp.execute()
    E = inp.results.non.gs_non_E
    print(" xmax = {:4.1f}, E = {:6.4f} Ha".format(xmax,E))

In order to run this example, save it to a file run.py and do

python run.py

Using the iDEA package in an ipython shell

As iDEA is a python package it can also be run from the interactive python shell (ipython). This is particularly useful as ipython has a convenient auto-complete feature. The following example can be run in an ipython shell line-by-line.

from iDEA.input import Input

# import parameters file
pm = Input.from_python_file('parameters.py')

# change parameters
pm.run.EXT = True
pm.run.NON = True
print(pm.run)

# run jobs
results = pm.execute()

# plot the relevant results
x = pm.space.grid
plt.plot(x, results.ext.gs_ext_den, label='exact')
plt.plot(x, results.non.gs_non_den, label='non-interacting')
plt.legend()
plt.show()